Everyone experiences the eerie sense of time passing and wonders where the past has gone. Some brush these speculations aside as mere indulgence amid life’s harsh practicalities. Others dare to press their queries. Does the past exist independently of the present? Or is it just a figment of our collective imagination? Thus they come face to face with one of the great conundrums of philosophy. My book, falling under three heads, Truth, Time, History. argues that the past has indeed no independent existence, but that the famous figures and events of history are not fictions, Henry V really was the victor of Agincourt, and the Visigoths really did sack Rome in 410 AD .

Part I: Truth is about the battles between philosophers who believe the past is real and those who believe it unreal.. This ancient debate, belonging both to theory of knowledge and metaphysics, puzzled St Augustine way back in the 11th Century. As he recounts, no sooner had he reasoned that if the past exists, it must exist somewhere, than another thought stymied the first: “wherever the past is, if it is there as past, then it no longer is”. The most distinguished recent philosopher to assert the past’s unreality was Michael Dummett, for whom it is an application of a more general theory according to which truth is not about how the world objectively is, nor therefore transcends our knowledge, but is a function of present evidence, and so changes with changing evidence. He challenges his realist opponents, who reject this as preposterous, to say how we can then know anything about the past since we can never go back. The realist point to the “truth value links” holding between differently tensed statements made at different times. If I am now giving a talk on truth in Oxford, then I also know now, without awaiting evidence available at this later date, that it will be true in a year’s time that I was giving this talk a year ago. Nothing can surely change that!  Dummett replies that this is merely to apply a rule for transforming a present tensed statement, made now, into a past tensed one, envisaged as being made in the future. What makes anyone suppose that thereby they have secured ingress into the past? What they are representing as the past isn’t the past at all, just the present – Sophie giving this talk now. We cannot assume that the transformation, sanctioned now by the truth value link, will hold in the future, when we actually reach it, or hence give us now access to a past, such as it will be revealed then. Part I, after much discussion, pronounces this debate inconclusive.

Part II: Time, pursuing a different tack, argues that we cannot believe both that the past exists, and, as we all ordinarily do, that things persist through change. Consider a green leaf turning brown in autumn You will agree: it has to be the same leaf that was once green that is now brown. For if it were a different leaf – a brown one – it could not be said to have lost its greenness since it never was green. But the great 18th Century philosopher, Leibniz stipulates – surely incontrovertibly – that a thing cannot be the same as itself and yet have divergent properties from itself. So it seems it has to be the same leaf in order to lose its greenness, but being the same leaf means precisely retaining its greenness: a contradiction. So what you may say: philosophers’ worry is not ours: obviously things persist within limits, depending on their kind. But, are you saying, it will be replied, that it requires less than identity to persist? These are deep and difficult questions. It seems clear however that identity is an atemporal relation that cannot accommodate the unidirectionality of time’s arrow. It is eventually concluded that the contradiction cannot be resolved so long as the leaf’s having been green has an equal reality with the leaf’s being brown, and that this provides grounds for demoting the past, and maintaining that only present things exist.

Part III: History asks how this move helps with the contradiction.   Prominent presentists, claiming to resolve it, have held, among other things, that “past times” inhere abstractly in the present as stories. We reject this powerful theory as too cumbersome, and strained, in that its adherents, being realists about truth, yet denying that the past exists, have to postulate a new kind of metaphysical entity to serve as truth maker for historical statements. Our hybrid proposal, which accords a crucial role to historians, affirms realism as regards presently existing objects, including historical texts – the tomes on library shelves – but suspends it as regards the non-existent past which is their subject matter. We are  influenced by the philosophers of history, Collingwood and Oakeshott, for whom, being idealists, not realists, historical truth is a function of the coherence of historical interpretation, not of anything existing in the past. Historical texts are governed internally by coherentist principles. “The victor of Agincourt was Henry V” is, on our view, not about Agincourt or Henry V, but about the historical text in which it figures. When realists protest: ‘What nonsense! It is about a real king, and a real battle. History is not fiction! Otherwise historians could make up anything they liked!’ we reply: they cannot: historians are held to rigorous standards in interpreting evidence. How could it help anyway to postulate that shadowy past which no historian has ever been able directly to access in order to check a single conclusion? . Historians, it is finally suggested, by their intellectual powers and narrative technique, fulfil a crucial role in the construction of the historical past, and perhaps the past more generally..

Prominent presentists, claiming to resolve this contradiction,  hold that “past times” inhere abstractly in the present as stories.   This  ingenious theory is however cumbersome and strained.  Its adherents, being realists about truth, yet denying that the past exists, are obliged to  postulate a new kind of metaphysical entity to serve as truth maker for historical statements. My  hybrid proposal, which accords a crucial role to historians, affirms realism as regards presently existing objects, including historical texts – the tomes on library shelves – but suspends it as regards the non-existent past which is their subject matter. I am  influenced by the philosophers of history, Collingwood and Oakeshott, for whom, being idealists, not realists, historical truth is a function of the coherence of historical interpretation, not of anything existing in the past. Historical texts are governed internally by coherentist principles. “The victor of Agincourt was Henry V” is, on my view, not about Agincourt or Henry V, but about the historical text in which it figures. When realists protest: ‘What nonsense!   It is about a real king, and a real battle. History is not fiction! Otherwise historians could make up anything they liked!’  I reply: they cannot: historians are held to rigorous standards in interpreting evidence. How could it help anyway to postulate that shadowy past which no historian has ever been able directly to access in order to check a single conclusion? . Historians, it is finally suggested, by their intellectual powers and narrative technique, fulfil a crucial role in the creation and construction of the historical past, and perhaps the past more generally.

A  more indirect, but in my view,  more sure, route to the conclusion that the past is unreal  takes off  from  the notion of identity over time.  I  argue that it is impossible to believe  both that the past exists, and also, as we all ordinarily do, that things persist through change.  Consider a green leaf turning brown in autumn.   You will agree: it has to be the same leaf that was once green that is now brown. For if it were a different leaf – a brown one – it could not be said to have lost its greenness since it never was green. But the great 18th Century philosopher, Leibniz stipulates – surely incontrovertibly – that a thing cannot be the same as itself and yet have divergent properties from itself. So it seems it has to be the same leaf in order to lose its greenness, but being the same leaf means precisely retaining its greenness: a contradiction.  So what you may say: philosophers’ worry is not ours: obviously things persist within limits, depending on their kind. But, are you saying, we will reply, that it requires less than identity to persist?   These are deep and difficult questions. It seems clear however that identity is an atemporal relation that cannot accommodate the unidirectionality of time’s arrow. It is eventually concluded that the contradiction cannot be resolved so long as the leaf’s having been green has an equal reality with the leaf’s being brown, and that this provides grounds for demoting the past, and maintaining that only present things exist.

Everyone experiences the eerie sense of time passing and wonders where the past has gone. Some brush these speculations aside as mere indulgence amid life’s harsh practicalities.      Others dare to press their queries.   Does the past exist independently of the present? Or is it just a figment of our collective imagination?      Thus they come face to face with one of the great conundrums of philosophy. My book, falling under three heads, Truth, Time, History. argues that the past has indeed no independent existence, and yet  that the famous figures and events of history are not fictions: Henry V  was the victor of Agincourt, and the Visigoths  did sack Rome in 410 AD .

St. Augustine  was already puzzling, in the fourth century,  over  whether  the past was  real or unreal.  As he famously recounts, in his Confessions, no sooner had he reasoned that if the past exists, it must exist somewhere, than another thought stymied the first: “wherever the past is, if it is there as past, then it no longer is”.    The most distinguished philosopher of recent times to assert the past’s unreality is Michael Dummett, for whom it is an application of a more general theory according to which truth is not about how the world objectively is, nor therefore transcends our knowledge, but is a function of present evidence, and so changes with changing evidence.  To realists, dismissing  this as utterly  preposterous, he replies: how then can we  know anything about the past since we can never go back.  His opponents point to the “truth value links” holding between differently tensed statements made at different times.   If I am now giving a talk on truth in Oxford, I  know now, without awaiting evidence at this later date, that it will be true in a year’s time that I was giving this talk a year ago. Nothing can surely change that!   Dummett   replies that this is merely to apply a rule for transforming a present tensed statement, made now, into a past tensed one, envisaged as being made in the future. What makes anyone suppose that thereby they have secured ingress into the past? What they are representing as the past isn’t the past at all, just the present – Sophie giving this talk now. We cannot assume that the transformation, sanctioned now by the truth value link, will hold in the future, when we actually reach it, or hence give us now access to a past, such as it will only be revealed then.   This debate, I argue however,  is finally inconclusive.