R.G. Collingwood – An Early Archaeological Theorist?

I had been long enough in my laboratory; I wanted to exchange it for my study. It was time to begin arranging and publishing the lessons which all this archaeological and historical work had taught me about the philosophy of history.


Introduction
It is often said that Collingwood found time to pursue two quite separate careers: as a philosopher; and as an historian (and archaeologist). But in fact he pursued three separate careers, for, as we will see, in Collingwood's view archaeology and history are distinct disciplines. Since Collingwood's time, the distinction between archaeology and history has become blurred and so this point is rarely appreciated. The philosopher Jan van der Dussen is exceptional in his understanding of the matter: The first thing to note is the fact that Collingwood was active in the fields of both archaeology and history…it is obvious that these subjects cannot be sharply divided, especially in Collingwood's own practice. The subjects are not, however, the same, and if one mentions only one of them in connection with Collingwood's activities this can be misleading. (Van der Dussen 1981: 201-02) This paper argues that although the lines that Collingwood drew between archaeology, history and philosophy are different from the lines that most would draw today-and although it would be inaccurate to term Collingwood an archaeological theorist-nonetheless, the seeds of archaeological theory are to be found in his work and in the work of certain of his contemporaries and predecessors. than was possible within the limits of a short article, how my conception of historical research was developing. (Collingwood 1939: 120-21) He was keen that Roman Britain should demonstrate what use the historian might make of archaeology. In his view the distinction between history and archaeology is such that, although archaeology might be termed 'the methodology of history' (Collingwood 1993: 492), there is more to history than archaeology. Thus, the gathering and classifying of data that occurs in archaeology is, for Collingwood, an 'essential preliminary' (Collingwood 1999: 52) to the writing of history; but, aside from archaeology, history has another methodology-a methodology that is unique, 'which lays down universal canons for dealing with all kinds of sources and constructing any kind of narrative about any subject' (Collingwood 1993: 492). This 'pure methodology' is the philosophy of history, 'a science dealing with the universal and necessary characteristics of all historical thinking whatever, and distinguishing history from other forms of thought' (Collingwood 1993: 492). Collingwood (the philosopher) argues that the distinguishing characteristic of history is that it reconstructs the rational chain of thought that is expressed in an action. When the historian asks what happened, he is asking what thought occurred. For the historian the question of 'what happened?' is synonymous with 'why did it happen?' Collingwood does not make the point explicit, but it may be inferred that the same does not hold for the archaeologist. The archaeologist is less interested in the question of why certain events occurred.
Collingwood does not discuss the distinction between archaeology and history at any great length for although he accords greater scope to archaeology than many of his contemporaries (recognising, for example, that archaeologists deal with both written and unwritten evidence), his views on this particular matter were not otherwise untypical of his time. He regards archaeology as a form of natural science that serves history.
Collingwood states explicitly that his next book, The Archaeology of Roman Britain, is a work of archaeology: 'it is strictly… a handbook of archaeology, not a history' (Collingwood 1930: vi). The book falls within his description of the archaeological sciences as a "means to criticise sources and extract history from them" (in Van der Dussen 1981: 213).
The distinction, as Collingwood saw it, between history and archaeology may be obscured by the fact that, in his philosophical writings, he sometimes uses the term 'critical history' as a synonym for 'archaeology.' As the following passage makes clear, archaeology ('critical history') is conceptually different from history in that it is only the latter that reconstructs an agent's reasoning. By contrast: Critical history classifies its sources into groups, and then subdivides these groups, and then subdivides these groups, framing rules for the manipulation of the various subdivisions. Taken as a whole, this technique is an abstract or classificatory science, which has no general name, unless that of archaeology is used for it, and is subdivided into numerous departmental sciences. (in Van der Dussen 1981: 213) The Archaeology of Roman Britain fulfils this classificatory role exactly.
In Collingwood's view, archaeology (or critical history) consists in a straightforward process of the rejection and acceptance of sources, on grounds of their anachronistic or authentic nature, but the reconstruction of an agent's reasoning is not a necessary part of this process (Collingwood 1993: 258-9;1999: 14). It is only in that the archaeologist serves the historian, and so is principally interested in artefacts that served human purposes, that his work is distinguishable from that of, for example, the palaeontologist. Granted the finished product of the archaeological process is history of a basic form; but upon reaching this form it is handed to the historian for further analysis.
Unlike Roman Britain, The Archaeology of Roman Britain was intended not so much to nurture the interests of the casual reader as to solve specific problems of chronology, such as might be encountered by archaeologists on site. It gathered and classified data that had hitherto been scattered across a growing number of specialist publications. A new edition, published in 1969, incorporated the revisions of Collingwood's pupil, I.A. Richmond. It was published four years after Richmond's death in 1965 and was credited to both R.G. Collingwood and I.A. Richmond. In reviewing the 1969 edition, Wheeler has this to say about the original: I saw much of Robin when he was writing the book. For him, with his essentially broad habit of thought, it seemed at first a trifle off-beat that he should immerse himself in so much museum-like detail at the expense of his more specific interests in the diverse fields of professional metaphysics, the wide movement and meaning of history, and the intriguing puzzledom of epigraphy, to say nothing of his periodical journeyings up and down Roman Britain. I remember wondering whether he was not really fashioning this textbook as a sort of imposed personal discipline. In many ways he was by nature a survivor from that long and distinguished line of amateurs-in that fullest and best sense of the term-who for three centuries or more had sustained the study of British antiquities as an inevitable and engrossing duty of the educated mind. The true end of this line had been marked by the brief but powerful transit of Pitt Rivers…But Pitt Rivers had died in 1900, and now it was 1930 and time for a receptive mind like Collingwood's to get his details right, along modern lines. In the process, he might help other aspirants into new, often ill-sorted, knowledge; but that admirable motive, I felt sure at the time, was incidental to a primary mission to organize his own thinking. (in Van der Dussen 1981: 220) If Wheeler's hunch, that the book was 'incidental to a primary mission to organize his own thinking,' is accurate, then we might expect to find some repercussions of this work in Collingwood's philosophical work. And, indeed, it seems that we do: for two years later, in an Essay on Philosophical Method, Collingwood describes systems of non-overlapping classification as the defining characteristic of empirical science. The examples that later commentators usually mention in this respect are botany and geology. Both organise the objects of their study into mutually exclusive classes; but so too does Collingwood (the archaeologist) in The Archaeology of Roman Britain. The exact route by which Collingwood's experience of classification in archaeology fed into his philosophy has yet to be traced-Collingwood would of course have encountered problems of classification in archaeology for many years prior to writing The Archaeology of Roman Britain-but undoubtedly when Collingwood describes systems of non-overlapping classification in An Essay on Philosophical Method he would have had recent first-hand experience of formulating such systems.
Collingwood's next volume on Roman Britain, Roman Britain and the English Settlements, was seen by both Collingwood and his peers as his historical magnum opus. It was written with J.N.L. Myres but, as Collingwood was keen to point out: 'This work is not a work of collaboration. It consists of two independent studies of two distinct, though interlocking subjects' (Collingwood and Myres 1936: v). Collingwood wrote on Roman Britain, and its immediate aftermath: and his half of the book was intended to supersede earlier editions of Roman Britain. This was the last major work of history published under Collingwood's name during his lifetime. His remaining years were to be devoted to philosophy for: I had been long enough in my laboratory; I wanted to exchange it for my study. It was time to begin arranging and publishing the lessons which all this archaeological and historical work had taught me about the philosophy of history. (Collingwood 1939: 121) By this time, Collingwood had formulated what he believed to be a distinguishing characteristic of historical explanation, namely, re-enactment. This most misunderstood of all Collingwood's ideas amounts to the claim that when the historian offers an explanation of an historical agent's actions he offers a reconstruction, albeit at a rudimentary level, of that agent's chain of reasoning. As mentioned previously, the historian reconstructs the rational chain of thought that is expressed in an action. Collingwood argues that this is a characteristic of historians' explanations that is absent from explanations in natural science.
Collingwood advanced this claim not as an archaeological theorist but as a philosopher of history. The difference is that archaeological theory is ultimately empirically driven-moved by what works in practice, and what does not; by contrast the philosopher of history's primary interest is in the logical structure of explanation. That is not to say that Collingwood, as a philosopher, was entirely uninterested in what worked in practice and what did not. As a philosopher of history, Collingwood sought out and studied what were commonly accepted as the best practices of his day. His work held up a mirror to these practices, but the prescriptive element to his work is recessive: as a philosopher, it was not his foremost intention to advise on how to dig or how to write history.
Collingwood was well aware that thus far archaeology and history had developed in a satisfactory manner without any help from philosophers. In particular what he termed the 'Baconian revolution' in history (Collingwood 1939: 115), which had occurred in the study of history around the turn of the century, had occurred beneath the notice of philosophers. This was to the discredit of philosophy but as Collingwood was keen to point out the disregard of philosophers had had no obviously detrimental effects upon the study of history and archaeology. In fact, it was with a certain amount of glee that he praised his mentor, the decidedly non-philosophical Francis Haverfield, as an intellectual revolutionary (Collingwood 1939: 83). Nonetheless, he did not think that this situation reflected well on philosophy; and, as a philosopher, he aimed to remedy it.
Finally, we must consider Collingwood's posthumous The Roman Inscriptions of Britain (1965), completed by R.P. Wright. This consists of an illustrated catalogue of over 2,000 Roman inscriptions. As a catalogue, rather than a work of reconstructed reasoning it falls, in Collingwood's terms, within the field of archaeology rather than history. This work had been initiated by Mommsen and taken over by Collingwood, from Haverfield, at the latter's death in 1919. In 1938 Collingwood took on R.P. Wright, the new lecturer in Roman History at Durham, as junior editor; and it was under the names of Collingwood and Wright that the work was finally published in 1965.
Curiously Collingwood himself, in the course of writing Roman Britain and the English Settlements, seems to have grown somewhat disillusioned with what the study of inscriptions might achieve. He recognised their bias towards frontier armies at the expense of all else (Collingwood and Myres 1936: 464); and this prompted the following remarks in his autobiography: As documents, inscriptions tell you less, under critical scrutiny, than literary texts; regarded as relics, they tell you less than archaeological material proper. And on the questions which I particularly wanted to ask, it happened that inscriptions threw hardly any light. (Collingwood 1939: 146) These remarks are revelatory of Collingwood the historian. The questions that he particularly wanted to ask concerned the lower strata of society. He was interested in the extent to which the lower strata of society were affected by Roman mores. This interest was long standing: it can arguably be traced all the way back to his work on his father's 'now classical excavation of a Romano-British village' (Collingwood, 1939, 80) at Ewe Close in 1908. He was the first to realise the need to examine our own use of the concept 'Romanisation' (Collingwood 1939: 140): he realised that the story of Roman rule could not be told solely from a top-down perspective. He was interested in the effects of the conquest upon those conquered. In all of these respects, Collingwood seems a very modern figure.
What is missing from his armoury is a sensibility towards attitudes to the past within the past. There is nothing in Collingwood's archaeology, history or philosophy that would preclude such awareness, but he is more focused upon the onward sweep of history. This has to do with his motivation as an historian of Roman Britain, aside from his philosophical interest in history and archaeology.
As judged against the norms of today, his motivation is unusual. He has some admiration for Roman achievements-The New Leviathan (1942) attests to his admiration of Roman lawbut he makes it very clear that he has no natural aesthetic affinity for the culture of the Roman Empire, and he is even less keen on the particular culture of Romano-Britain. This is most strikingly revealed in his chapter on 'Art' in Roman Britain and the English Settlements in which declares: on any Romano-British site the impression that constantly haunts the archaeologist, like a bad smell or a stickiness on the fingers, is that of an ugliness which pervades the place like a London fog: not merely the common vulgar ugliness of the Roman empire, but a blundering, stupid ugliness that cannot even rise to the level of that vulgarity. (Collingwood and Myers 1936: 250) We also know from a letter to his friend Guido de Ruggiero that he did not visit Rome until 1927(Collingwood 1927) (by which time he had already written Roman Britain, the standard textbook on the subject).
So what then was his motivation? Intellectually, in Collingwood's view, it just would not do for the period of Roman rule to remain an isolated period in Britain's history. He believed that archaeology might make this period more understandable and better integrated into Britain's entire history. Within the entire history of Britain the history of Roman Britain would then hold the equivalent place to the story of a man's schooldays within the story of his life (Collingwood and Myres 1936: vi).
As in the story of a man's schooldays, in which a large part of their significance is seen in what they made possible at a later date, so, for Collingwood, the significance of the Roman Empire lies in what it made possible. It created the conditions in which Christianity might eventually flourish (Collingwood 1932: 149). When we talk of Collingwood's motivation as an historian and archaeologist we should not forget that he is a Christian historian (see Collingwood 1993: 49-52).
But to return to our central theme: it is important to establish Collingwood's views of the relationship between archaeology, history and philosophy in order to guard against anachronistic readings of his work (the distinction that Collingwood draws between archaeology and history is, for example, overlooked by Hodder and Hutson (Hodder and Hutson 2003: 145-52) and, until recently, by myself (Leach 2009)). Collingwood would, I believe, have endorsed Collingwood Bruce's dictum that archaeology is 'the handmaid of history' (Collingwood Bruce, J. 1857: 7). However, he did not endorse the oftenaccompanying belief that history deals with written sources and that archaeology deals with unwritten (Collingwood 1939: 133). Written sources, for example memorial stones, may be 'richer' than unwritten sources in that they already incorporate a rudimentary proto-historical understanding of a past event, but their study is not invariably more profitable. The same questions can reasonably be asked of unwritten sources as of written, with the same or greater reasonable expectation of an answer. It so happened that in the case of Roman inscriptions Collingwood found that, of the questions that he was most interested, inscriptions had less to tell him than unwritten sources.
Neither should we assume that he equates archaeology with data and history with interpretation. The line between data and interpretation is drawn in all his major archaeological and historical works. Admittedly, in the interests of fluency, the line is feint in Roman Britain and Roman Britain and the English Settlements but it is still there. It is by virtue of this line that his various interpretations have either been incorporated into more recent histories, or rejected, or placed 'on hold.' For all that he wrote as though he himself had witnessed Caesar's invasion, Collingwood would not have expected otherwise. He deliberately left his work in a state that it might be re-interpreted.
The distinction between Collingwood's works of archaeology and history seems to be rather that the former were written primarily in answer to the question 'what?' and the latter were written primarily in answer to the question 'why?' This way of distinguishing between archaeology and history has fallen a long way from fashion since Collingwood's time. The Russian archaeological theorist Leo S. Klejn upholds this distinction (see Klejn 2001), but few others (Klejn upholds it in the wake of his experience of Soviet archaeology in which there was a strong tendency for it to be deemed only necessary to know 'why?' in order to answer the question of 'what happened?'). Thus, although they are out of step with most current archaeological theory, Collingwood's beliefs with respect to the relationship between archaeology and history are still upheld, albeit in perhaps unexpected places.

From the Study to the Laboratory
Up to this point we have concentrated upon the lessons that Collingwood took from his laboratory to his study; but, in his autobiography, Collingwood also wrote of the lessons that he took from his study to his laboratory; and in this ambition alone he will at once seem to provide more familiar company to contemporary archaeological theorists.
From his study to his laboratory Collingwood took the following three principles of 'historical thinking' (Collingwood 1939: 121): 1) Long practice in excavation had taught me that one condition-indeed the most important condition-of success was that the person responsible for any piece of digging, however small and however large, should know exactly why he was doing it. He must first of all decide what he wants to find out, and then decide what kind of digging will show it to him. This was the central principle of my 'logic of question and answer' as applied to archaeology. (Collingwood 1939: 121) 2) A second principle was that, since history proper is the history of thought, there are no mere 'events' in history: what is miscalled an 'event' is really an action, and expresses some thought (intention, purpose) of its agent; the historian's business is therefore to identify this thought. (Collingwood 1939: 127-28) 3) A third principle was that no historical problem should be studied without studying what I called its second-order history; that is, the history of historical thought about it. (Collingwood 1939: 132) Taking these principles in reverse order: the third principle is now extensively followed with respect to specific archaeological and historical problems. Collingwood himself provides the example of his rediscovery, in old archaeological papers, of Roman fortifications on the Cumbrian coastline in just the places where his hypothesis had suggested that they might be found (Collingwood 1939: 129). However, archaeologists' knowledge of the history of archaeology remains somewhat erratic. For example, the claim that archaeology can ask the same questions of written evidence as of unwritten evidence, with the same reasonable expectation of an answer, is still advanced as though it were a new realisation-even though David George Hogarth made the same claim already in 1899 (Hogarth 1899: v). And, for example, the claim that archaeology is 'the handmaid of history' is often conflated with the claim that archaeology deals with non-written evidence and history deals with written evidence-but, as we have seen, these claims do not necessarily coincide.
The second of Collingwood's principles is rooted in Collingwood's analysis of the presuppositions underlying historical explanation. For reasons of space I shall say no more about this principle here. It is a claim that has been the subject of intensive scrutiny within the philosophy of history.
It is the first of Collingwood's principles with which I am here concerned. This principle is also to be found in Flinders Petrie's Methods and Aims in Archaeology (1904) in the following maxim: The old saying that a man finds what he is looking for in a subject is too true; or, if he has not enough insight to ensure finding what he looks for, it is at least sadly true that he does not find anything that he does not look for. (Petrie 1904: 49) In other words, it is a necessary (but not sufficient) condition in archaeological investigation to have some question in mind to which one seeks an answer. Collingwood was not the discoverer of this principle-and never claimed to be-but, with others such as his friends Wheeler and Crawford, he was one of its most energetic propagators.
The juxtaposition of these principles may give the impression that Collingwood does not draw a distinct line between archaeology and history, but this is not the case. As has been seen, a clear distinction is implied in the preface to The Archaeology of Roman Britain (Collingwood 1930: vi) and elsewhere (Collingwood 1993: 492;1999: 52). Archaeology follows the 'logic of question and answer'-arguably more accurately termed the method of question and answeronly because it serves history, and in particular, it serves to answer the questions that the historian formulates. Of course the archaeologist and the historian may be embodied in a single person, but the conceptual distinction remains. With the exception of Leo Klejn few contemporary archaeologists would follow Collingwood in drawing this distinction. Moreover, most contemporary philosophers are keener to join their own discipline to that of natural science rather than to regard all other disciplines from a position of isolated detachment.
Nonetheless, setting these debates to one side, I would argue that it is the first of these principles that is the logical foundation of archaeological theory. Sometimes the initial question must be broad and sometimes reassessment and new or additional questions are required, but these are refinements: no one would now exchange the first of Collingwood's principles for that of Mr Micawber. Collingwood, writing in 1938 told his archaeological readers that they would be astonished to the extent to which the archaeological world of the 1920s was ignorant of this principle (Collingwood 1939: 126). Since the 1930s it has become more deeply embedded yet and it is in large measure the acceptance of this principle that has enabled the development of archaeological theory. Its acceptance implies the acceptance, in practice, of some element of 'why?' in archaeology-in that the archaeologist must have an answer in mind to the question, 'why are you doing this piece of work?' Nonetheless, despite this indispensable element, Collingwood would still claim that archaeology is primarily motivated towards the question of 'what?' whereas it is history that is primarily motivated towards the question of 'why?' Thus, the lines that Collingwood drew between these disciplines are different from those that most would draw today. Yet, for all that, the seeds of contemporary archaeological theory are to be found in his work and the work of certain of his contemporaries and predecessors.
Philosophy Programme, Keele University