Figure 127

Recall as a function of perspective at the time of retrieval. Based on data from Anderson and Pichert (1978).

alternative perspective, and then to recall the story again. On the second recall, participants recalled more information that was important only to the second perspective or schema than they had done on the first recall (see Figure 12.7).

These findings support the notion of schema-driven retrieval. Further support comes from the participants' own accounts (Anderson & Pichert, 1978, p. 10):

When he gave me the homebuyer perspective, I remembered the end of the story, you know, about the leak in the roof. The first time through I knew there was an ending, but I couldn't remember what it was. But it just popped into my mind when I thought about the story from the homebuyer perspective.

Script-pointer-plus tag hypothesis

The script-pointer-plus-tag hypothesis was put forward by Schank and Abelson (1977). It represents a development of some of Bartlett's ideas, and consists of a number of assumptions about memory for script-or schema-based stories:

• Information from the story is combined with information from the underlying script or schema in memory.

• Actions in a story are either typical (consistent with the underlying script or schema) or atypical (inconsistent with the underlying script).

• Information about atypical actions is tagged individually to the underlying script.

• Recognition memory will be better for atypical than for typical actions, because typical actions present in the story are hard to discriminate from typical actions absent from the story.

• Initial recall for atypical actions should be better than for typical actions, because they are tagged individually in memory.

• Recall for atypical actions at long retention intervals should be worse than for typical actions, because recall increasingly relies on the underlying script or schema.

The evidence generally supports the prediction that recognition memory for atypical actions is better than for typical ones at all retention intervals (Davidson, 1994). However, the findings with respect to recall are more inconsistent. Davidson (1994) shed light on these inconsistencies. She used routine atypical actions that were irrelevant to the story and atypical actions that interrupted the story. For example, in a story about going to the cinema, "Sarah mentions to Sam that the screen is big" belongs to the former category and "Another couple, both of whom are very tall, sits in front of them and blocks their view" belongs to the latter category. Both kinds of atypical actions were better recalled than typical ones at a relatively short retention interval (1 hour), which is in line with prediction. After 1 week, however, the routine, irrelevant atypical actions were less well recalled than typical or script actions, whereas the interruptive atypical actions were better recalled than typical actions. As Davidson (1994, p. 772) concluded, "Part of the problem with existing schema is that they do not specify how different types of atypical actions will be recalled."


Our organised knowledge of the world is used in a systematic way to help text comprehension and recall. However, it has proved hard to identify the characteristics of schemas. More importantly, most versions of schema theory are sadly lacking in testability. If we are trying to explain text comprehension and memory in terms of the activation of certain schemas, then we really require independent evidence of the existence (and appropriate activation) of those schemas. However, such evidence is generally not available.

One of the major assumptions of schema theory is that top-down processes lead to the generation of numerous inferences during story comprehension. However, the evidence discussed earlier in the chapter (e.g., McKoon & Ratcliff, 1992) suggests that the number of inferences generated by the average reader is less than is implied by schema theory.

Finally, the reader may have noticed that memory rather than comprehension was the primary focus of the research discussed in this section. Real-world knowledge manifestly affects comprehension processes, but schema theory does not indicate in detail how this happens.

Kintsch and van Dijk's model

One of the most successful models of discourse processing was put forward by Kintsch and van Dijk (1978). There are two basic units of analysis within their model: the argument (the representation of the meaning of a word) and the proposition (the smallest unit of meaning to which we can assign a truth value; this is generally a phrase or clause). The text of a story is processed to form structures at two main levels:

• The micro-structure: the level at which the propositions extracted from the text are formed into a connected structure.

• The macro-structure: the level at which an edited version of the micro-structure (resembling the gist of the story) is formed.

Kintsch and van Dijk (1978) argued that the propositions extracted from a story are entered into a short-term working buffer of limited capacity similar to the working memory system proposed by Baddeley and Hitch (1974; see Chapter 6). Additional propositions are formed from bridging inferences, and added to those formed directly from the text itself. When the buffer contains a number of propositions, the reader tries to link them together in a coherent way. More specifically, propositions that share an argument (i.e., two words having the same meaning) are linked. Linking of propositions occurs only within the buffer, and thus is limited by the capacity of short-term memory. There is a processing cycle: at regular intervals, the buffer is emptied of everything but a few key propositions. Propositions are retained in the buffer if they are highlevel or central in the evolving structure of the story, or if they were presented recently.

The macro-structure of a story combines schematic information with an abbreviated version of the microstructure. Various rules are applied to the propositions of the micro-structure:

• Deletion: any proposition not required to interpret a later proposition is deleted.

• Generalisation: a sequence of propositions may be replaced by a more general proposition.

• Construction: a sequence of propositions may be replaced by a single proposition that is a necessary consequence of the sequence.

Memory for the text depends on both the micro-structure and the macro-structure. Higher-level or more central propositions are remembered better than low-level propositions, because they are held longer in the working buffer and are more likely to be included in the macro-structure. This prediction has been confirmed several times (e.g., Kintsch et al., 1975).


Evidence for the importance of propositions was obtained by Kintsch and Keenan (1973). They manipulated the number of propositions in sentences and paragraphs, but equated the number of words. An example of a sentence with four propositions is: "Romulus, the legendary founder of Rome, took the women of the Sabine by force", whereas the following sentence contains eight propositions: "Cleopatra's downfall lay in her foolish trust of the fickle political figures of the Roman world". The reading time increased by about one second for each additional proposition.

Ratcliff and McKoon (1978) provided good evidence for the existence of propositions. They presented sentences (e.g., "The mausoleum that enshrined the czar overlooked the square"). This was followed by a recognition test in which participants had to decide whether test words had been presented before. For the example given, the test word "square" was recognised faster when the preceding test word was from the same proposition (e.g., "mausoleum") than when it was closer in the sentence but from a different proposition (e.g., "czar").

McKoon and Ratcliff (1980) presented participants with a paragraph. This was followed by tests of recognition memory, with participants deciding whether the ideas contained in sentences had been presented in the paragraph. The response times to perform this recognition-memory task were speeded up when a sentence was preceded by another sentence from the paragraph. This speeding up or priming effect was determined more by closeness of the two sentences within the propositional structure of the micro-structure than by closeness in the text, thus providing evidence for the reality of the micro-structure.

Kintsch (1974) distinguished between effects of the micro-structure and of the macro-structure on memory for text using a verification task. Participants decided whether explicit and implicit inferences were consistent with a text that they had either just read or had read about 15 minutes earlier. Explicitly stated propositions were verified faster than implicitly stated propositions on the immediate test, but there was no difference in verification time between the two kinds of propositions after 15 minutes. According to the theory, explicit propositions are better represented than implicit propositions in the micro-structure, but both are equally well represented in the macrostructure. Information in the micro-structure is much more available immediately than after a delay, and this explains the different pattern of results at the two time intervals.

The construction-integration model. Adapted from Kintsch (1992).

Stop Anxiety Attacks

Stop Anxiety Attacks

Here's How You Could End Anxiety and Panic Attacks For Good Prevent Anxiety in Your Golden Years Without Harmful Prescription Drugs. If You Give Me 15 minutes, I Will Show You a Breakthrough That Will Change The Way You Think About Anxiety and Panic Attacks Forever! If you are still suffering because your doctor can't help you, here's some great news...!

Get My Free Ebook

Post a comment