There for representing images and verbal is

There are several theories attempting to explain the way in which people represent knowledge. Three main theories explaining mental representation and imagery are Dual Coding theory, Propositional Theory and Mental Models. Each theory suggests ideas that share similarities and vary to a certain extent. They look at the analogous and symbolic ways in which data can be represented in the mind. The aim of the essay is to explain main ideas of the mental representation models with evidence for each theory and compare similarities and dissimilarities in their concepts and recreation of information for problem solving.


One of the first theories is Dual-coding theory that is based on a notion that information is represented by two perceptual codes: visual and verbal (Paivio, 1969). The visual code is for representing images and verbal is for words as well as other concepts and symbols. Visual information is regarded to be analogous as it is a similar representation to the real-life object, and verbal information is used for more symbolic representation, which means that representations do not perceptually replicate objects or concepts. The proof of the two distinct codes was presented by Paivio(1973) when participants were shown an arrangement of images or words and asked to freely recall presented verbal and visual information tended to recall more images; and, when asked to recall items in order recalled more information that was presented verbally. This difference in performance infers that information is processed differently depending on the type of what is being processed and represented. 


Another experiment that looked at animation that was either preceded, accompanied or without verbal explanation of the process of what was shown, in that case, the work of a pump (Mayer & Anderson, 1991). The study hypothesised three outcomes looking at single-code, dual-code and integrated hypothesis. When given a problem-solving task, participants who viewed animation with verbal narration scored higher than the ones that listened to the narration before or had none. One obvious explanation for better performance is that participants had a verbal and concrete image of the pumping process, so had access from two codes, which is in agreement with Dual Coding Theory. The second explanation of better performance is that there was a link between mental representations of two codes when having concurrent audible and visual information.


Propositional theory, unlike Dual Coding Theory, is not based majorly on primary sensory information and is viewed as an alternative point of view on mental representation. Representation is seen more of an epiphenomenon and has no perceptual features as Dual-coding theory has, instead, it implies an underlying meaning and relationships of an object or a concept (Pylyshyn, 1973). It is believed that images and verbal information are stored as propositions and propositions are then used when recreating those types of representations. Pylyshyn argued that representations are not processed further as images can be perceptually processed. The evidence for the existence of propositions was shown using ambiguous images (Chambers & Reisberg, 1985). Participants were shown images that could have represented two different objects and asked to state the percept and the second possible image. One of the explanations was that propositions are stronger than visual codes depending on the context. Though the image did not infer one image over the other, participants had difficulties when trying to observe the second possible image. Which means that the mental images are not analogous to what is perceived which is theorised in Dual Coding Theory.  An earlier experiment (Carmichael, 1932) looked at the reproduction of visual imagery that was slightly ambiguous. Two different word lists were presented to two groups labeling the same ambiguous image that could fit both descriptions. It was found that visual representation could be altered with verbal labels and suggested that mental representations are not a parallel replica of perception. Clark (1972) added that mental representations need to act as a single code when he looked at how participants compared sentences with images. He assumed that when seeing printed word and an image of the same object an individual would use a single mental representation code to draw a link, as there are no common modes to infer the same meaning. Thus, people use a type of propositions that contain underlying meanings of what they interpreted verbally or visually, unlike double code way of representing incoming sensations.


Some of the evidence comes from experiments with visually impaired participants and haptic maps (Kerr,1983). Similar to the study by Kosslyn (1978) about map scanning, Kerr presented blind and sighted participants with a board with topographical features of a map and asked them to create a mental image of the map. The blind group though responded slower, but when asked to scan distances, spent more time on longer distances and less time for shorter distances. Haptic responses revealed other modalities of mental representations and since blind participants did not have a conventional visual experience as sighted participants they should have had a mental model that represented geographical distances. Since Dual Coding theory looks at visual imagery, looking and proving at other modalities of representation could serve as an evidence of mental models formed, also, Kerr (1983) showed that representations of a map could be non-analogous to the visual image.  Propositions of spatial relations of a map couldn’t be formed by a visual sense in blind participants, which also agrees with the possibility of mental model forming.


Cognitive psychologists proposed combining the views on how people form mental representations. Mental Models theory is a type of mixture of Dual Coding and Propositional Theory that was proposed by Johnson-Laird (1994). Mental Models are seen as more of implicit knowledge modules developed by individuals to interact and represent objects and concepts. In comparison to Dual coding theory and to some extent propositions, Mental Models are dependent on personal experience and can be faulty and influence mental representations. Mani and Johnson-Laird (1982) observed how participants represented verbal information on spatial location and links of four objects. Half of the descriptions were certain and specified and the second half was partly uncertain. Participants later had to decide whether or not the diagrams presented to them matched verbal information. Mani and Johnson-Laird theorised that if a mental representation is ruled by propositions that both kinds of descriptions of object relation in space should be equally remembered. That is because both types of information would lie under the same representational code. On the other hand, a certain description would be remembered better if there is a mental spatial group because uncertain descriptions would need two varying representations. Participants showed better results with determinate information, however, when asked to recall descriptions word for word indeterminate information was better recalled. This trend was explained by the notion that propositional representation is formed of the sentences that include enough information to remember direct verbal input. The second step would be to create an internal model of spatial relations the basis of the proposition. More mental processing is involved in creating a mental model, which makes it more memorable, but does not contain specific details. Unlike mental models, propositions are less memorable but contain more specified information.


When comparing the theories, it can be seen that Dual Coding Theory and Propositional Theory are alternatives to each other. One looks at mental representation as being created via two codes (Paivio, 1969, 1973) and the other focuses on how mental representations work similarly to abstract propositions (Pylyshyn, 1973). Propositions are the underlying meaning of an object or a concept and their relationship to other phenomena. Mental models theory is a mixture of propositions, mental imagery and mental models (Johnson-Liard, 1993). Mental models are dependent on individual’s experience and are an implicit knowledge that is used to explain and comprehend various experiences, objects, and concepts. Whereas Dual Coding theory is more perceptually based and perceptually similar to a represented phenomenon, Mental Models and propositions to some extent are more dependent on prior experience and abstract.


If one looks at how mental representations are transformed or recreated to solve different problems, three theories offer different explanations. If Dual Coding theory is based on analog and symbolic type of encoding (Paivio, 1969), it can be said that modification of mental representations is similar to real life objects. Mental Rotation task (Shepard & Metzler, 1971) showed that people mentally rotate objects similar to the way they would do it in real life. The degree of rotation positively correlated with the time taken to match an original object with a rotated object. This implies that mental representation is analog to the percept that could be further evidence for visual codes used for mental imagery.  As for the propositions, they are thought to affect the recreations of knowledge-based as stored concepts can distort a visual image. In the experiment by Friedman (2009), he looked at the mapping of cities in North America and how participants tended to represent stored geographical knowledge. It was shown that certain beliefs concerning categories of cities affected how people estimated the location of cities, which may be seen as stored propositions of relations of cities to each other and underlying modality relations. There are many categories influencing mental representation, which may be seen as propositions. Like Propositions, Mental Models Theory is said to be combining different schemas; and, one should be able to use heuristics and alter the stored knowledge blocks to be able to solve problems (Merrill, 2000). When comprehending a text, Glenberg (1987) argued that mental models affect what is remembered better and that mental models are an updatable phenomenon. Unlike Dual Coding theory, mental models are said to be encoding the meaning of the piece of information like text, not the visual properties of verbal or textual information. It was found that mental representation of a text resembled how events of a story were spatially related and located in the text.



To conclude, there are three main theories that attempt to explain the way that knowledge is mentally represented. These theories have influenced a lot of research and each affected development of other theories. Three theories discussed in the essay share some similarities as well as differences. Thus, Dual Coding theory is focused more on a direct representation of the perceptual features of an object or a phenomenon through two codes and Mental Models and Propositions are more abstract as well as providing information on relations to other things and underlying meaning of objects and concepts. The latter two use perceptual data to retrieve or recreate images, sensations, and other additional information. When recalling data, dual coding theory is said to be able to recall visual information better when recalled randomly, whereas verbal information is recalled better when is recreated in order is was perceived in.  Propositional theory focuses more on how images or verbal information can be reproduced but the help of propositions that explain underlying relations and meaning of objects and concepts. As for mental models, some believe knowledge is believed to be recreated in a similar way to propositional theory. Propositions serve as a basis of a mental model and can facilitate recall.