As mentioned most research has been carried out on the comprehension of written texts. Sometimes the research uses specially constructed reading material and therefore more control can be taken over the variables which affect comprehension and sometimes published articles or books are used which of course provide greater ecological validity to the studies. Of course the former of the two methods gives the experimenter the ability to manipulate the reading material in a systematic way. Graesser et al. (1997) proposed the solution to the question of which method it better. “Discourse psychologists are on solid footing when a hypothesis is confirmed in a sample of naturalistic texts in addition to properly controlled textoids [experimenter-generated texts].”
It would be impossible to comprehend discourse without access to stored knowledge; the process of inference demonstrates the vital role played by that stored knowledge. Inference is described by Schank (1976) as “the core of the understanding process”. We make inferences, which help our understanding whenever we hear a sentence and the process is so natural that we sometimes do not know we are doing it. Bridging inferences are made to establish understanding between the current part of the text and the proceeding text and elaborative inferences embellish or add to details of the text. It is widely accepted that most readers draw bridging inferences from text, which are very important for comprehension however the subject of exactly how many non-essential or elaborative inferences are drawn is a matter of more debate. Anaphora is an example of the simplest form of bridging inference; a pronoun or noun has to be identified with a previously mentioned noun and using the appropriate anaphoric inference can do this. For example “John gave Fred his toy”. In this sentence the inference that the toy belongs to John is clearly understood by us because we have learnt an automatic understanding of this rule over time. In order to establish this anaphoric inference the distance between the noun and the pronoun has to be considered. The closer the pronoun is to the noun to which it is referring the easier it is to establish the inference; this is known as the distance effect. Clifton and Ferreira (1987) discovered that when measuring the time it takes for a participant to read a sentence the distance between the pronoun and the noun had no effect on reading time but reading time was dependent on whether the relevant noun was still the topic of discourse. Therefore the distance effect is found only because a greater distance reduces the probability that the noun to which it refers is still the topic of discourse when the pronoun is presented.
The greatest strength of the minimalist hypothesis is that it distinguishes between those inferences, which are automatic, and those, which are not. Constructionalist theorists argue that any inferences we need to draw from a sentence are done so automatically however it is dubious to assume there would be no ambiguity in deciding what information needs to be encoded in order for our understanding.
With story processing the whole matter becomes more complicated. When we are asked to recall a story we will relay the major events and themes of the story but leave out the specific details. We do not simply relay the story word for word as it was told to us or read by us. What we do remember of the story is usually its meaning and we are highly selective about our description. Gomulicki (1956) found that when participants were given a précis of a story and were asked to recall the same story from memory and they found it very difficult to tell the two apart. In other words story memory resembles a précis and people focus mainly on important information.
Most if not all stories have a basic structure and it is argued that all stories share common elements at a very general and abstract level. This whole concept led to the notion of story grammar, which is a set of rules from which any given story can be generated. Thorndyke (1977) suggests that there is a hierarchical structure in any story with the major categories of themes, plot, setting and resolution are at the top. He tested this hierarchy by presenting stories to participants, one had the theme at the start one at the end and one omitted the theme altogether. Recall of the story was better when the theme was at the start but it was better to have the theme at the end then not at all. Meyer and McConkie (1973) found that an event low down in the story hierarchy was better recalled if the event immediately above it in the hierarchy is remembered. The story grammar approach is not flawless however. Harley (1995) identifies one reason. “There is no agreement on story structure: virtually every story grammatician has proposed a different grammar.” Another reason is the fact that the story grammar does not reveal much about the process involved in story comprehension.
The next theory is the schema theory; it involves the use of chunks of our in built knowledge of the world, which are used as a framework to interpret stories and events. Scripts and frames are relatively specific kinds of schemas. Scripts involve events and consequences of events: Schank and Ableson (1977) referred to a restaurant script, which contains the usual information about the process of going to and eating in a restaurant. Frames on the other hand deal with knowledge about the properties of objects and locations. Schemas are important when processing language as they contain the knowledge we need in order to understand what we read and hear. Schemas allow us to form expectations and make the world a more predictable place than it would otherwise. If an event did not fit in to our particular schema of things we would take action in order to solve the problem. Some people have varied schemas of events however on the whole we all share generally accepted schemas, which we are probably not even aware of. Bransford and Johnson (1972) used a short instruction passage, which was given to participants. The participants who did not know what the instructions were for rated the passage incomprehensible as it did not fit into any of their pre-existing schemas and those who knew the title (“washing clothes”) found recall much easier and they recalled 5.8 idea units as opposed to 2.8 if the title was not given.
Bartlett (1932) is probably one of the chief psychologists involved in the schema theory. According to him memory is affected not only by the presented story but also by the participants store of relevant prior knowledge in the form of schemas. Bartlett came up with the idea of presenting people with stories, which provided people with a conflict between what was being presented to them and their prior knowledge. For example a story from a different culture may produce on recall a distortion of facts twisted to fit in with their own schemas. Bartlett’s research backed up his theory. He found in particular that a substantial proportions of errors made on recall were in order to make the story more conventionally English. He called this finding rationalisation, which refers to the type of recall error made. Bartlett assumed that memory about a precise object is forgotten over time whereas schemas are not. Therefore according to this theory the longer information is retained the more rationalisation errors our made using schemas. There are problems with Bartlett’s research. He did not give participants very specific instructions and therefore some of the distortions may have been due to conscious guessing rather than deficient memory. This point was backed up by Gauld and Stephenson (1967) when they found that when participants were given instructions stressing the importance of accurate recall half the errors usually obtained were eliminated.
