The fact that my pupils and others cheerfully write to Lara Croft as if to a real person, can be attributed to various factors. Apart from any other considerations, as far as my pupils were concerned, a teacher told them to do it. However, the fact that pupils did it cheerfully and enthusiastically. suggests a relationship between the child and what Sherry Turkle refers to as an "object to think with".

I do not propose to suggest that pupils cannot tell the difference between Lara Croft and a real person. However, for them the stars of East Enders are not ‘real people’ either. They differentiate between the people they meet in real life and the people they only see on the screen, whether it is TV or video game.

When working with Logo programs, pupils have to look at life from the ‘point of view of the turtle". For example, they need to be able to imaginatively conceptualise the idea of turning right in order to go left if the turtle is facing down the screen. I am interested in the extent to which pupils are doing the same thing in relation to Lara Croft in Tomb Raider.

Far from talking about the "effects" of computer games on pupils, Sherry Turkle characterises the computer as a Rorschach. Not only do different pupils choose different games but more importantly they react to the same game in very different ways and use it for different purposes. The Rorschach concept suggests that the pupil is projecting meaning onto the game rather than being a passive recipient of other people’s meanings. (Turkle, 1984,5)

The attitudes and values which children are projecting on to the game may well be accounted for by the "effects" of other influences. Pupils are subject to peer or parental pressures and they are in receipt of messages from other media.

Nevertheless pupils do seem to be seeking some kind of mastery of the game. It is the effect of this self-assertion by the pupil and the consequent relationship with the screen character of Lara Croft which concerns me

Observing pupils using Logo, Sherry Turkle characterised two kinds of mastery. The "soft mastery" of the pupil who wants to develop an interactive relationship with the program and the ‘hard mastery’ of the pupil who wants to impose his or her will on the program. Her research suggested that there were far more female "soft masters" and are more male "hard masters" but that the personality of the pupil expressed in their style of mastery was not solely dependent on gender (Turkle, 1984,103).

She later developed this idea further. In Life on the Screen, she refers to the modernist approach of trying to get as close to the reality of the computer’s machine language as possible and she contrasts this with the postmodern approach which is more comfortable with layering and simulation - effectively working in ignorance of the underlying mechanism.

The example of the Macintosh computer which places a simulated desktop between the user and the operating system is used. The user is still involved in manipulating this environment but the modernist imperative "analyse and you will know" is absent. The system involves a much more exploratory attitude on the part of the user. The users learn to negotiate rather than analyse. The user does not have access to the ‘lower levels’ of the system which are closer to the hardware and need not be aware of them.(Turkle 1995, 43)

Computer programming is not as prevalent a use of computers as it was when computing was a specialised subject taught to a few pupils. For most pupils, programming takes the form of "customising options." This could mean writing complex Excel macros which are close to the traditional concept of computer programming. It could also mean something as simple as changing the resolution of the Tomb Raider screen.

Nevertheless, the modernist aesthetic which Sherry Turkle recognised can be seen in the popularity of ‘cheats’ which seek to bend the program to the pupil’s will.

Initially assembly-code programmers produced cheats for games. By examining the source code of the game they could produce programs which could give the player infinite lives, invulnerability or skip to a new level by altering the contents of registers which contained the player’s number of lives, level of stamina or level number in binary code. This programming typically took place outside the corporations which produced the games.

Another kind of cheat is information, for example maps of the levels of the game. Notice how the very use of the word ‘map’ suggests the description of a real space which the player is exploring. In the earlier text-only games the ‘map’ would take the form of a set of directions for completing the level.

The more recent development of cheats clearly required the collaboration of the manufacturers. Typically cheats now consist of lists of key presses required to achieve the advantages which the assembly language programmer only acquired with considerable skill. These lists are published openly in magazines, some of which are produced directly by the manufacturers of the games.

In any case it is physically very difficult for hackers to probe and uncover the secrets of console games and the built-in cheats are the only opportunity for the player to achieve some sense of ‘hard mastery’ over the game.

On the other hand there are still cheats available on the Internet for PC games which are unlikely to have been created by the manufacturers. The nude raider patch is just an extreme example of such a cheat. It is presumably not sanctioned by Eidos (the manufacturers) and is widely excoriated by the fans of the game in the Internet usenet groups and on the nude-raider-free web sites.

One year 9 pupil who contributed to the Letters to Lara project wrote:

"Lara I know I will not have you as a girlfriend but on my computer I can control you. if you were my girlfriend I would know which buttons to press."

Under cover of jest, this pupil seems to be suggesting an attitude of hard mastery and apparently a rather worrying attitude towards girls in real life. The "knowing which buttons to press" metaphor suggests a highly manipulative attitude. He told me he had heard it from a TV program and thought it referred to erogenous zones (he didn’t actually use the words). As far as I can tell, his relationships with other pupils are not characterised by attempts to manipulate or bully them.

Notice also that he is hovering around the edges of treating Lara as a real person, "I will not have you as a girlfriend," because she is famous and rich or because she does not really exist? If that is the case then the rather dangerous and manipulative image which his message suggests is modified.. I think it is accurate to say that although the pupil is being facetious, he is also using the figure of Lara Croft as a transitional object. He can clearly differentiate between how he can relate to this transitional object and how he can relate to his peers

Transitional objects are, for example, objects to which small children can safely remain attached when they transfer their attention between concentration on the mother and encountering the outside world. (Winnicott, 1971) For older children the computer or games console can become a transitional object because it can be seen as "belonging simultaneously to the self and to the outside world" (Turkle, 1984, 218).

The Wiggins survey found self-esteem scores were not significantly related

to time spent in watching television, playing arcade or home video games, or in programming home computers. External locus of control seemed related to time spent in playing arcade games or in home computer programming. However, pupils who participated extensively in playing arcade games or in working with the home computer seemed more externally locused than did other subjects. (Wiggins, 1985)

This finding is very interesting despite the fact that in video game terms it is very old. Of course a correlation does not prove a causal link. It may be the case that pupils who feel they are more externally locused are in fact using the video games to create a world in which they are more in control. In terms of transitional objects, it does not make sense to talk about regular video game players being externally locused. If the game is an extension of the self then the locus is not "external".

The concept of the transitional object is helpful in understanding the way children treat computers. The view of computers as ‘alive’ or ‘conscious’ is encouraged by TV science fiction in which the aggressive, quirky or irritable computers function as characters. An example of this attitude is seen when Weizenbaum (1984) found that people, including psychotherapists, treated his cybershrink ELIZA program as if it were far more intelligent than it really was.

Another example of the anthropopathy of computers is that Nass et al found that people were even attached to individual computers to the extent that they would not complain to a computer about its performance but preferred to complain to another computer about it. They did not want to be rude to their ‘own’ computer.(Nass, 1994)

Erdman et al reported a growing tendency for people to trust the computer with the highly "human" task of psychotherapy. People were more likely to be candid with computers. The wide knowledge base to which a computer had reference and the changes in the perception of the role of psychotherapy made computers more likely candidates for this kind of work. (Erdman et al,1985)

The concept of cyber-pets captures one aspect of the computer as perceived by users. The cyber-pet is totally dependent on the user. This relationship might give the user a feeling of power which compensates for the powerlessness they may see in their situation as children in relation to adults. The cyber-pet stands in the same relation to the child as the child might stand to the adult in some relationships.

In my seventh year class, there are a variety of attitudes towards cyberpets. Some revel in the responsibility which the child has for the welfare of the cyber-pet and the ostensible feelings of guilt and distress when one dies. These feelings are often denied by older, usually male pupils who profess to take pleasure in killing off cyber-pets or may express their mastery by telling younger children how to bring the cyber-pet back to life. Both attitudes may be expressed by the same child at different ages.

Very few pupils will go on to develop the intense symbiotic relationship between the hacker and the computer described by Weizenbaum (Weizenbaum, 1984) but the cyberpet phenomenon puts the concept of computer "personality" in a different light. Human beings fear the computer as a controlling intelligence (a good deal of science fiction makes use of the 2001 concept of a computer turning on its creators) but nobody fears a smart pet. Likewise, despite her formidable arsenal, none of the children in the Letters to Lara project expressed fear of Lara Croft.

The cyberpet phenomenon also highlights the companionship which children get from computers. The companionship which children get from pets is something which parents do not regard as threatening or dangerous. When children get a similar companionship from books, from the time-shifted presence of the author-as-friend, few parents regard this as dangerous. However, when children gain the same companionship from the computer, from the time-shifted presence of the programmer-as-friend, this is regarded as a step away from humanity. In addition there are many parents who might think Lara Croft is a highly unsuitable companion for their child.

The relationship between the pupil and Lara Croft can be seen as much closer to the relationship with a cyber-pet than the relationship with a real life girl.