‘EVERY HISTORIAN SHOULD write an autobiography,’ wrote the historian AJP Taylor, introducing his own, A Personal History (1983). ‘The experience teaches us to distrust our sources which are often autobiographical.’ In other words, it teaches us humility about the basis of the story we are telling – not that humility is a quality particularly evident in Taylor’s work. The main danger, he thought, was that the historian-autobiographer might exaggerate his successes or, more likely, his failures and humiliations. But that was easily corrected: ‘the experienced historian ends by striking out the more fantastical episodes, even if they happen to be true.’ This should produce the uninteresting personal history that Taylor, as he assures us, set out to write.

What Taylor called autobiography most theorists would call memoir, the difference (as expounded by Karl J Weintraub, in a classic article in Critical Inquiry in 1975) being that memoir is a record of external fact while autobiography is a reflection on the inward realm of experience, an attempt to find the meaning of a life. A historian can easily tell the two genres apart by the rule of thumb that if something lends itself to being used as a source, it’s a memoir, and if it proves strangely recalcitrant, it’s autobiography.

The American writer Mary McCarthy wrote both, starting with autobiography. Memories of a Catholic Girlhood (1957), written in her mid-thirties, was an effort to give a meaning to the pain of her harsh upbringing in a great-aunt’s house after her parents’ death, and the resentment she felt about it. Returning to the subject thirty years later, in the memoir How I Grew (1987), she retreated from the intense emotion of the autobiography and submitted her memories to a rigorous critical review (‘On reflection, I see that I have been exaggerating’; ‘But stop! That cannot be true…’). The imaginative writer, in other words, had become a historian of the strictest, more fact-oriented kind.

When I read How I Grew in the late 1980s, I immediately thought of my mother, Dorothy Fitzpatrick, a lifelong crusader against inaccurate generalisations (a phrase she might have found redundant), who in her eighties was turning her critical fire against her own memories: ‘I see now that I got it all wrong,’ she would say with grim satisfaction, correcting yet another error in her previous accounts of an academic success or her relations with her father. The accounts were unwritten, for the process of writing, in which some generalisation is almost unavoidable, was agony for her. She did agree to be interviewed by a young researcher about my father’s civil liberties work, but when he gave her the transcript, which I’m sure was accurate, she filled the margins with anxious doubts, disclaimers and corrections of her own corrections.

My father, the radical historian Brian Fitzpatrick, was a big-picture man for whom writing was an easy and pleasant thing, especially if he was upending somebody else’s conventional wisdom. As a historian, I seem to be more my father’s daughter than my mother’s, though not in every respect. His was a politically engaged history, telling a story of Australian development that hinged on class conflict of Labour and Capital, with Labour as the progressive force. My histories of Soviet Russia, in contrast, are politically disengaged (at least in my view: some of my American colleagues might disagree); and I think I always had a stronger sense of the relativity of truth than my father, though I could be wrong about that. My histories were written in the mode that Soviet critics used to label ‘so-called objective’ history – a small step up from being a ‘bourgeois falsifier’, as they called many Western Sovietologists, but still necessarily prejudiced. Finding myself in the United States in the 1970s, I had my own problems as an outsider in that Cold War Sovietological community. Like Nikolai Sukhanov in his Notes on the Revolution (1922), I took a certain pride in being ‘alien, indifferent and polemically disposed’.

HOW IS A historian to write the history of her own life? My mother spent her last years carefully unpicking the version she had fashioned in early and middle adult life. My father put his resentful memories of childhood and youth into an unpublished novel, ‘The Colonials’, written in his mid-twenties, and then tried to forget about his family altogether, to the point where he claimed not to be able to remember his sisters’names. I took a different approach in the book that ultimately became My Father’s Daughter: Memories of an Australian Childhood (MUP, 2010). I planned to write the history of myself and the family I grew up in, using my familiar tools as a historian, but with a lighter touch than my scholarly works and a more literary style. The light touch was important to my conception of the book: it was to be a memoir, detached and (I hoped) entertaining, not an autobiographical probing of the depths. The focus would be on my father, and certain aspects of my own personal life would be omitted as irrelevant.

I tried out the approach in a short piece on my father in the London Review of Books (8 February 2007), and was pleased to find that readers responded, even though most of them didn’t know who my father was. That was an important discovery: by writing, you can bring the dead to life. At first it was just my father (who died in 1965) who was to be the object of resurrection; my more reticent mother joined the picture later. Locally famous, if not notorious, in the 1950s for his left-wing politics and drinking, as well as for the provocative Australian economic histories written outside academia, Brian seemed to have fallen into obscurity after a brief posthumous flourishing in the Whitlam era. I thought he would like to be brought back to life, even warts-and-all.

True to my training in the University of Melbourne History Department, as well as the example of my father and mother, I am a historian who lays great stress on primary sources. In the case of the memoirs, I was my own chief primary source, so the first thing to do was get down my memories in as unadulterated form as possible. That meant not exposing them to contamination from other sources (a lesson I learnt from interviewing old Bolsheviks, whose memories in some cases were not only highly polished by previous retelling but also, infuriatingly, amplified and corrected in light of their later reading of scholarly books, including mine). I sat in my temporary office at the University of Sydney, typing away, chapter after chapter. It was much easier and quicker than drafting a scholarly work because I never had to look anything up.

Once I’d finished producing the primary source (that is, the first draft), the next step was checking. That’s when the problems started. I took informal oral histories from all the old friends and family members I could find. I consulted my father’s and mother’s papers in the National Library. When I got back to Chicago, I went through my own letters and document files, organising them like an archivist for the purpose. I was alarmed by the results. First, there were the factual discrepancies and lacunae. Witnesses disagreed about key dates and events. People would tell me stories of events of that I couldn’t remember, even when prompted. I would tell them about events in which, according to me, they had participated, and they would sometimes remember vaguely, sometimes not at all, and occasionally deny outright that anything like that could have happened. The vague recollections often crystallised later into firm ones – but in at least one case I later discovered that the memory I had implanted in a friend was actually false. It’s true that I found a few witnesses whose memories seemed solid and reliable, like my matter-of-fact uncle Alan Douglas, who knew my immediate family well without being a part of it, and whose corrections of my own flights of memory I always took as gospel. But these were few and far between.

As a historian, I had always thought I was properly sceptical about the reliability of memory, but my scepticism came to seem puny and inadequate to the dimensions of the problem. I observed numerous instances where my memory, or someone else’s, had been enlisted to illustrate some cherished generalisation, being significantly though unconsciously altered in the process – the event had been moved to another time or place, with different people involved – to improve the fit. In other cases, memory had been adjusted simply to make better sense. An example arrived by email just recently, when an American scholar wrote to me about his memory of meeting me and Katerina Clark (Manning’s daughter) together in Moscow circa 1964. It makes perfect sense, two Australian Soviet scholars known to be friends who both went to Moscow in the 1960s, of course they would have been there together – but actually we weren’t: Katerina got there first and was gone by the time I arrived. So many discrepancies in people’s memories, and so much invisible self-editing, not to mention self-censoring! I started to see our brains functioning like word processing on the computer: in order to tell a story, you bring up a memory file, revise it in the telling – and then save the revised file, overwriting the original one. And, as we all know to our cost, when you overwrite a computer file, it’s gone.

There were problems with the written documentation, too. To be sure, there were some interesting discoveries, notably in my father’s letters to my mother and my mother’s letters to Eleanor Dark when I was a baby. But I was more impressed by what the written record left out than what it contained. Some things that I considered very important about my early life turned out to be absent from the surviving documents, and even from the oral histories. People really close to me turned out to have left almost no documentary trace, while letters from casual acquaintances cluttered up my old files. My parents corresponded with each other only when my father was away doing research in the Mitchell Library in Sydney, which for financial reasons happened increasingly rarely in my middle childhood.

Yet, comparatively speaking, the written record available to me was abundant, with Don Watson’s fine biography of my father (Brian Fitzpatrick: A Radical Life, 1979), Brian’s autobiographical novel, his papers and my mother’s preserved in a well-organised archive in the National Library, and the two sides of my correspondence with the family after I went to Oxford in 1964 also preserved. The dismaying thought came to me, though, that if someone else were to write the history of my life on the basis of all the sources except my memory, they would (in my mother’s habitual phrase) get it all wrong.

Given all this, I could have ended up writing a postmodern self-reflexive book about my search as a historian for the truth about myself, offering, Rashomon-style, all the different versions of the truth I had come up with. Had I been in residence at the University of Chicago at the time, the very air of the place might have impelled me to do so. But, as it happened, I was temporarily in a different, less rigorously postmodern environment, the Wissenschaftskolleg in Berlin. The more I thought about the story of my early life that seemed to be emerging, the more I thought it was a story with its own drama, and I ought to tell it straight. I changed my mind about leaving out ‘irrelevant’ personal stuff, since it now seemed to me that, unless I put it in, it would be distorting not only my own experience but even my relationship with my father.

I became much less interested, as I embarked on the second draft, in the question of factual correctness, since it was now an emotional truth rather than a literal or historical one that I was after. That doesn’t mean that I lost my interest in getting the facts right – I was quite punctilious in letting the reader know where my memories might be wrong – but it didn’t seem of prime importance any more. As I got into the fifth and sixth chapters, with my father’s death and another death that I had initially meant to leave out, I noticed that writing certain passages made me cry, not only the first time but also on fourth or fifth revision, and even on reading the copy-edited manuscript and page proofs later. It occurred to me that I probably wanted the reader to cry, too – not the usual aim when you’re writing history. I realised that something was happening to my book in the process of writing, and after a while I realised what it was: the memoir was turning into autobiography.

THE STAKES IN autobiography are different from those in a memoir, which at least shares the historian’s concern with catching the specificity of a time and place. In autobiography, as in literature, you’re trying to capture the specificity of experience, to convey a personal emotional truth rather than a strictly historical one. You may, if you choose, assume a pose of detachment in an autobiography but, paradoxically, the autobiography is likely to work only if the reader sees through it.

Thus Edmund Gosse in his autobiographical Father and Son (1907) writes as a wryly detached observer, but the whole power of the work comes from the reader’s perception that this is just a way of keeping intense emotion in reasonable check. I mention Gosse because, when I had recently to teach the book, which I had quite forgotten, to an undergraduate class in Chicago, it suddenly came back to me how important it had been for me, and for my father who introduced me to it, when I was young. It’s a son’s account of the misery his father inflicted on him, his anger, and the hopeless pity he felt for his father’s failures and humiliations at the same time. My own father’s favourite book, obviously speaking to his deepest concerns about his father: how could I have forgotten it? It was part of my sentimental education at my father’s hands to learn, via Gosse, that love in a family context – I thought then, in any context – was an unstable compound of pity and resentment. It now occurs to me that it was perhaps from my father that I learnt how to write My Father’s Daughter.

It felt good to have produced an autobiography, something of a catharsis. But when the suggestion came to write another book about my past, I instinctively headed back to the genre of memoir, not autobiography. There are things in my adult life (after My Father’s Daughter) that are still fraught with painful emotion, and it was very clear to me that I wasn’t going to write about them. What I felt like writing about was Cold War Moscow in the 1960s, a wonderfully interesting and eventful period of my life of which I have basically happy memories. It was the time of my first major research, my first archives and my first really foreign country (England turned out not to qualify). The multiple inconveniences and discomforts that seemed to be built into the Soviet system, the remarkable disconnect between rhetorical claims and everyday reality, and the pervasive preoccupation with secrecy and spying, were all an object of wonder to the little group of foreign students from ‘capitalist countries’, as was the warmth of the friendship of Russians, though we had been assured on the British side that friendship would be impossible. Since the Soviets regarded their archives as an almost sacred repository of state secrets, an archive-seeking historian was likely to find herself on the front line of the Cold War.

As with My Father’s Daughter, I did a trial run for the memoir in the London Review of Books, under the title ‘A Spy in the Archives’ (2 December 2010), and all sorts of other people who had once felt like spies in a foreign countries responded. I’m fairly confident that this time my memoir won’t morph into autobiography. But at this stage of the process last time, I thought the same thing. As of now, I think of myself as a much more reliable witness on my experiences in 1960s Moscow than I was about my childhood and youth, almost at the level of my uncle Alan. But we’ll see how well that holds up when I get to the actual writing.

IT REMAINS TO try to answer the tricky question I asked at the beginning: can a historian write a history of herself in the same way that she might write of something outside herself, like the Russian Revolution? Pace AJP Taylor, I doubt it. Primary sources are something for historians to use critically, and that’s very hard to bring off if you are your own primary source. Taylor himself didn’t bring it off, and by the end of his book you start to feel that the claim to have done so is one of his teases. Surely as a historian he would not have used only one primary source (his own memory) without checking it against others. If he had been looking coolly at a body of evidence not generated by his own memories and coloured by his own emotions, he would probably not have left in, unexplained, flatly contradictory data (pertaining, for example, to his attitude to his mother). Nor, I believe, would he have dropped a whole failed marriage (the equivalent of an Imperial reign!) from his historical narrative of, say, the Hapsburg Empire, even if somebody (the ex-wife, in this case) asked him to.

I tried to go one better than Taylor in My Father’s Daughter by treating my memory as a primary source and checking it against other primary sources. But the results, though interesting to me personally, were not crucial to the overall story of the book, since I decided ultimately to tell my story and, in the process, made the genre shift from memoir to autobiography. To be sure, I scrupulously alerted readers – perhaps as much in unrecognised tribute to my mother as from my profession as a historian – to the places where my story might be ‘wrong’. But I’m not sure that in the end those corrections made the book any more like a work of history. If you’re writing as a historian, you don’t usually include all the details of your weighing of the evidence; you just tell the story that the bulk of the evidence seems to support. Weighing the evidence from primary sources is assumed to be part of the job description of a historian: it’s only when the author’s primary source is her own memory that extensive commentary seems called for.

Looking back at My Father’s Daughter, I think my corrections function less as an approximation of historical method than as an indicator of anxiety: the author’s unreliability as a narrator may unwittingly jeopardise the autobiographical project of self-understanding. Philippe Lejeune, the pre-eminent theorist of life writing, has written about the unspoken ‘autobiographical pact’ whereby the author guarantees to readers to tell the truth about herself as she knows it. That’s emotional truth he means, not factual. When Mary McCarthy produced the corrected version of her life (with the apparent intent of undermining the grand narrative of her earlier autobiography), the second version didn’t seem truer to the reader – just less interesting, because the vivid lines of the autobiographical self-portrait were blurred. Reading the second book, you understand that she thinks she is approaching her past with the detachment of a historian, but you don’t really believe it.

Before I tried writing about myself, I think I assumed that, because of (quasi-)objectivity and broader reach, historians can give a superior account of the past to memoirists or autobiographers. That unexamined assumption coexisted with a lifelong preference for reading memoirs and autobiographies rather than histories (in our family of historians, my mother, the non-writer, was the only devoted reader of history books). When I’ve finished writing a historical work, I generally have the sense that I ‘got it right’, or as right as is possible given the state of the evidence. After finishing My Father’s Daughter, my private sense of achievement was more exalted: I felt that, after all the agonising, what I had arrived at was ‘true’, even ‘the truth’. That was a nice feeling for a congenital relativist. It took a little time for me to get back to normal, and to see that there might be more than one ‘true’ story to be told about my father and his daughter. But that doesn’t really matter. I still feel as if I upheld Lejeune’s autobiographical pact and told as much of the truth about myself as, at the time of writing, I was capable of knowing.