Memory and written storage                                                             

THEORETICAL ASPECTS

Types of memory

Most readers will be familiar with the experience of looking up a telephone number and then repeating it to themselves for the time it takes to sit down and dial the number. As luck would have it, this is invariably the occasion for somebody to ask a distracting question with the result that the number is forgotten and has to be looked up all over again. Equally familiar and irritating is when you need the same number twenty-four hours later and find that you are quite unable to remember it.

These experiences reflect the widely recognised view among psychologists that with verbal learning the ability to hold information over brief periods (usually up to thirty seconds in duration) demands fairly constant repetition, and any distraction or interruption is likely to severely impede that ability. Moreover, it has been established that our capacity for short term retention is remarkably consistent, and that most people experience some breakdown in retention as soon as the number of items or chunks of information exceeds seven.

This type of memory, known as short term memory, is clearly different from long term memory, which is our capacity for recall of information minutes, weeks and years after the original input. Furthermore the difference is not simply one of duration. Unlike short term memory which is limited in capacity, long term memory is seemingly inexhaustible and can accommodate any amount of new information. Not surprisingly this additional information can only be stored at a price; it is generally acknowledged that we need to work much harder to commit information to long term memory, and the type of repetition we described as being essential to short term retention may not be adequate for long term retention.

Some readers may feel uneasy about this last comment, as it would seem to contradict an experience we have all shared, namely the ability to remember certain information either by means of repetition, or with no conscious attempt to learn it at all. This certainly does happen, and the distinction between short term retention and long term retention is not always clear-cut. Information entering short term memory may pass quite effortlessly into long term- memory, and some learners may find repetition a very effective way of transferring information into long term memory.

  Organisation of the mental lexicon

  Our 'mental lexicon' is highly organised and efficient. Were storage of information haphazard, we would be forced to scan in a random fashion to retrieve words; this simply is not feasible when one considers the speed at which we need to recognise and recall. Furthermore, it is extremely improbable that we organise words in the brain as a dictionary does. Imagine you were trying to recall the word 'nozzle', for instance. It is unlikely that you would retrieve the word 'noxious¹ (which appears next to 'nozzle' in the LDCE) in place of the target word.

Principles in learning and teaching vocabulary

One way in which researchers investigate how the mental lexicon is organised is by comparing the speed at which people are able to recall items. It is generally accepted that if certain types of prompts can be answered more quickly than others, then this will reflect the lexical system. Freedman and Loftus (1971) asked testees to perform two different types of tasks: e.g.

1 Name a fruit that begins with a p.

2 Name a word beginning with p that is a fruit.

Testees were able to answer the first type of question more quickly than the second. This seems to indicate that 'fruits beginning with p' are categorised under the 'fruit' heading rather than under a 'words beginning with p' heading. Furthermore, experimenters discovered in subsequent tests that once testees had access to the 'fruit¹ category, they were able to find other fruits more quickly. This seems to provide further evidence that semantically related items are 'stored together'. Most researchers appear to agree that items are arranged in a series of associative networks. Forster (1976, 1979) put forward the theory that all items are organised in one large 'master file', and that there are a variety of 'peripheral access files' which contain information about spelling, phonology, syntax and meaning. Entries in the master file are also held to be cross-referenced in terms of meaning relatedness.

Some very interesting experiments carried out by Brown and Mc Neil Principles in learning and teaching vocabulary (1966) exemplify this point forcefully and give us clues about lexical organisation. The experimenters gave testees definitions of low frequency vocabulary items and asked them to name the item. One definition was, 'A navigational instrument used in measuring angular distances, especially the altitude of the sun, moon and stars at sea'. Some testees were able to supply the correct answer (which was 'sextant'), but the researchers were more interested in the testees who had the answer 'on the tip of their tongues'. Some gave the answer 'compass', which seemed to indicate that they had accessed the right semantic field but found the wrong item. Others had a very clear idea of the "shape' of the item, and were often able to say how many syllables it had, what the first letter was, etc. It seems, then, that these systems are interrelated; at a very basic level, there appears to be a phonological system, a system of meaning relations and a spelling system.

  We also have to consider other variables which affect storage. One important factor here is word frequency; items which occur most frequently are also easily recognised and retrieved. Imagine a pile of cards, each representing an item of vocabulary. In this system, the most frequently used items are 'at the top of the pile', and therefore easier to retrieve. Recency of use is another variable, and, to return to the analogy of the pile, one can imagine words more recently used being at the top. These variables are concerned with the use of items, but it is also impor tant to consider when items were first learnt. Imagine a pile of words organised chronologically: the words learnt on the first day of a course would be at one extreme and those most recently learnt at the other.

Clearly, native speakers do not acquire all their vocabulary in lexical sets, but rather acquire items in a haphazard, chronological fashion, generally in a fairly predictable order of frequency. However, native speakers have many years in which to build up a comprehensive lexicon, whereas foreign learners are limited in this respect. Exploiting our present knowledge of storage systems to the full should allow us to attempt to speed up the learning process and facilitate storage. This will be true whether we are trying to clarify associative networks, classify by categories or organise the vocabulary syllabus in a way which will assist the contribution of frequency and recency of use and other variables.

  Task 1. What are the practical implications of this?   

  Why do we forget?

In spite of the efficiency of these various organisational networks in the memory, we still suffer lapses when we are unable to remember something that we thought was well established in our long term memory. Why does this happen?

One theory of forgetting suggests that information stored in the memory falls into disuse unless it is activated fairly regularly. In other words, we need to practise and revise what we learn otherwise the new input will gradually fade in the memory and ultimately disappear. This is called the decay theory.

In opposition to this theory is the notion of cue-dependent forgetting, which asserts that information does in fact persist in the memory but we may be unable to recall it. In other words, the failure is one of retrieval rather than storage. Evidence for this theory resides in a number of experiments. In one of these, subjects were given lists of words to learn and then tested on their powers of recall. Later they were tested again, only this time they were given relevant information to facilitate recall. For example, if a list contained the words 'sofa¹, 'armchair¹ and 'wardrobe', the subjects would be given the superordinate 'furniture' as a cue to help them. These experiments showed that recall was considerably strengthened by appropriate retrieval cues, thus suggesting that the information was not permanently lost but only 'mislaid'.

In addition to the theories of decay and cue-dependent forgetting there is further evidence that any significant mental activity undertaken before or after periods of learning can also account for poor learning and retention. The activities undertaken prior to learning may have a detrimental effect on our ability to absorb new input, while activities undertaken after periods of learning can interfere with the effective consolidation and retention of new input. How long this interference persists is difficult to determine but the effects are likely to be most acute in the hours immediately preceding or following periods of learning. This contrasts with 'decay' which is obviously more significant in accounting for memory failure over a long period of time.

One final point about forgetting is the rate at which we forget. It is generally believed that of the information we forget, eighty per cent is lost within twenty-four hours of initial learning. This may help to explain why testing activities carried out the day after input may yield rather distressing results, while further testing activities carried out a week later appear quite satisfactory. This rate of forgetting clearly has implications for revision and recycling which will be discussed later.

As language teachers, our main concern is to ensure that what is taught will be permanently retained in long term memory, so it is clearly a matter of some importance that classroom activities take account of these various theories, and strive to combat decay and interference while developing and facilitating efficient retrieval systems.

TOP