An autobiographical account may be inaccurate in various ways. It may be intentionally or unintentionally inaccurate – call these two kinds of error lies and mistakes. It may be factually or psychologically inaccurate. It is clear, for example, that Rousseau was incorrect about some of his factual claims. However, he argues that he is completely accurate about his own internal life. We only have his word for this; we can only have his word for it; there are no independent witnesses to his psychological states. But it is possible that he was lying or mistaken about aspects of his inner life. Of factual inaccuracies, some can be independently investigated and some cannot. Things claimed to be done in private, without leaving evidence, may be irrefutable. Of psychological inaccuracies there are various kinds. Authors may represent the thoughts, opinions and inner life of earlier selves according to their current perspective’s whim. They may ridicule, insult and denigrate the memory of earlier selves, or justify, motivate and explain their current stance by citing thoughts and beliefs that earlier selves may or may not have had. They may tamper with their earlier selves’ values (including morals), for the purpose, for example, of throwing their current values into sharper relief.

I am not going to consider the morality of autobiographical inaccuracy – that was the paper I abandoned. Instead I am going to consider some motives for intentional inaccuracy – lying – that are defensible to varying degrees. I hold as indefensible many of the psychological inaccuracies listed in the previous paragraph. To misrepresent one’s former "soul" for the sake of one’s current "soul" seems to me cheap and tawdry. To a good autobiographer, a past self is an intriguing thing, to be understood and explained to others; using that self untruthfully, to argue for a position, nullifies the project.

Manipulation of the past self for the sake of the narrative is sometimes legitimate. There is obviously a long slope of accuracy down which a writer can slide to bring drama and intensity to their story. My guess is that few autobiographers go very far down that slope. I will consider the first few steps downwards that I suspect are fairly common.

Autobiographical lies may be used to compact a story. Unless the sense is changed, these are often trivial and completely justified. I suspect that all autobiographers (even Primo Levi, who I regard as the most reliable of those we have read during the course) tell such lies. Sometimes it just isn’t worth spelling out the differences between similar events; it is better to conflate them. In my autobiography, I have sometimes done this. The miserable party and Ian’s guts ache were two separate events in our German trip, eating toasted teacakes in the power cut is a composite. Mary McCarthy goes further in this direction. Explaining how in "The Figures in the Clock" she arranged events to make a "good story", she says, "It is hard to overcome this temptation if you are in the habit of writing fiction; one does it almost automatically" ([3] page pages 164-165). It is worth noting that Rousseau, Sartre and Woolf were also in the habit of writing fiction.

Dramatic touches are forgivable when they are only little lies. It is most likely to be near the culmination of a story, that a particular poetic ending is desired and the truth must give a little. The example that comes to mind is Mary McCarthy’s story of the tin butterfly ([3] page 80). Her horrific account of abuse is brought to an almost elegant conclusion by the revelation that Uncle Myers, the accuser, planted the evidence. The author herself questions the accuracy of the ending in her commentary on the chapter (pages 82-83), and explains how she might have become mistaken. Of course, we only have her word that it was a mistake and not a lie! But even a lie might be forgivable here. My autobiography has a few examples of this, none so elegant as McCarthy’s, simply little exaggerations to open up or close off a story. We didn’t rouse the whole camp at Coniston, I didn’t come last in the Sheffield public speaking contest (but I came close to it), the crossing of Breydon Water wasn’t that rough.

Many times an author may have forgotten parts of the story and so interpolates plausible events. Again McCarthy confesses to this, for example, in her chapter "C’est le Premier Pas Qui Coûte". Her interpolations there are not only plausible but well researched. Sometimes I don’t remember details of stories in my autobiographical vignettes, so I have bent the story to fit. I don’t, for example, remember if the industrial accident victim that I met lost his arm in a power press or another piece of machinery, but presses are what we were worried about at Makin’s.

An autobiographical lie may be introduced for the sake of the formal structure of the work. I cannot identify any examples from conventional literature, but they occur in my autobiography because of the peculiarly constrained nature of the design. The need to provide links between all my vignettes has led to some truth bending. I have used a link called "dream girl" between two stories talking about different girls (though they both had the same name). I suspect that I have imagined our form teacher picking people at random to read out their essays. It would not surprise me if it were true, but I think it is not. Nevertheless I included it to get a good link to a page about speaking. Again, I think these linkage problems could be corrected with careful revision.

Accelerating down the accuracy slope brings us to the practice of "fictionalizing" adventures to render a substantive truth elegantly, compactly or entertainingly. McCarthy’s story "The Blackguard" is, she says, "highly fictionalized" and "true in substance, but the details have been invented or guessed at" ([3] page 97). It is here we should begin to be sceptical, for how can we be sure where the fiction ends? In McCarthy’s case, she tells us, but presumably her original readers in Harper’s Bazaar were not so informed. The check against highly fictionalized accounts is the testimony of other witnesses. But sometimes this is not available. Consider, for example, Sartre’s Les Mots [11]. In Sartre’s case the book overflows with his stylistic virtuosity, giving his subject, the young Sartre, the restless energy and vitality that is his true substance. What seems to be highly fictionalized is not the facts, but the endless array of mental impressions that process through young Sartre’s mind, described in vivid detail. I therefore suspect Sartre of psychological inaccuracy, and this, of course, that can never be tested. However, Sartre’s portrait almost asks not to be taken too seriously, so his sin of fictionalization – if such it is – is forgivable.

This, I think, is about as far down the accuracy slope that one can go, and still render an autobiography. Of course, there are plenty of examples of an autobiographical stem to other kinds of literature – novels, poetry and even technical writing. But the addition of invented detail soon takes the work beyond the borders of autobiography. Finally, though, what about the omission of material? Does a work that claims to be autobiographical lie if it silently passes over vital information? Lejeune’s comment about models may be relevant here ([12] page 23): talking about the resemblance of a text to a life, he says that the resemblance can be found "in the negative mode – and at the level of the elements of the narrative – [where] the criterion of accuracy intervenes; [and] in the positive mode – and at the level of the whole of the narrative – [where] what we call fidelity intervenes. Accuracy involves information, fidelity meaning." Surely the meaning of a life will be lost if key material is omitted. What is left unsaid cannot be critiqued, except, again, by reference to other witnesses. So to consider this final question I must turn again to my own partial autobiography.

My autobiography has been sliced two ways – by broad subject and by year. Many works are not like this: authors make no attempt to cover themes or every year of their lives. It is immediately clear in my case that only a time slice of my life has been presented. This may not be a lie, because autobiographies must be temporally localized. Readers should simply be careful not to assume that the chosen ten years tell the main story about me. But by using a tabular format during design I have also, consciously, excluded important threads of my life. My mother and father were the most important factor in my first ten years. They still had a significant influence in my next ten years, yet I have said very little about them. Religion was a defining thread in my life during my teens – at the time, I would have said the most important thread – yet I have hardly mentioned it. Does this mean therefore that the hypertext autobiography given here lacks fidelity – that it is, in effect, a big lie?

I am not sure about this. A life has facets; turn it and they catch the light. You never see inside, but the reflections show you patterns, and unseen facets mean hidden patterns. A perfect autobiography would describe a myriad facets, yet there could still be a missing angle that might reveal the most. All autobiographies must omit information, and how is the author to know in the end where the meaning is? So I do not feel too guilty about leaving out information. I think it probably is impossible to write autobiography without a lie of this kind. While Lejeune might think it worse than simple lies of inaccuracy, I regard it as natural and warn readers of my autobiography: expect accuracy not fidelity!

[9] Terrell Neuage, "The Influence of the World Wide Web on Literature: Bibliography" http://www.geocities.com/SoHo/Lofts/5289/THESIS2.html