The English Parish Church as an Object and Category of Study

In the Middle Ages, and indeed until relatively recently, the parish was the basic unit of ecclesiastical and secular administration in England. Originally, it was a domain of episcopal governance, but in the High and later Middle Ages, parochia referred to an area of indeterminate size centred for official purposes on a village or town. This area was served by a priest appointed to oversee the religious affairs of its inhabitants. The priest had his administrative seat in the town or village church. Said church was often the largest building in a parish, and in most cases it was the only one of functional significance to all parishioners regardless of age, sex or wealth. With the exception of extreme unction, it was normally the only place in which parishioners could receive the sacraments and in whose environs they could be buried. By extension, it tended to become a focus of cultural activity for the laity. The parish also served Church and Crown alike as a unit of taxation. For these and other reasons, it forms a coherent and obvious basis for investigation, and has been used as such in county histories of England since the sixteenth century.¹

In the context of art and architectural history, the choice of the parish in the form of its church is more a matter of convenience than historical logic. At its most flexible, ‘the parish church’ has been employed by scholars simply to corral object or image categories that would otherwise seem unmanageably open-ended.² More commonly, the intention is to investigate something that is considered intrinsically parochial. For scholarly and popular authors alike, some concept of the parish church as a distinctive artistic domain is usually involved in this. The concept has proven durable and enabled some highly useful work, whose ultimate significance is its stake in decision-making about the future of buildings and their contents.³ It is nevertheless worth asking how this generic concept serves the broader aims of scholarship, particularly in light of some of the difficulties it entails. This question is the business of the current essay. It should be acknowledged at the outset that what follows is neither exhaustive nor surprising in its propositions. The ideas discussed here will be recognisable to most readers. The essay’s main purpose is to critique something of basic relevance to this volume on the assumption that any self-respecting study should cross-examine the validity of its terms of reference, and further that doing so can produce healthier premises.

With this in mind, the proposal that the parish church as a category is by nature ambiguous, and has been (for the positivist at least) unhelpfully sentimentalised, is to play devil’s advocate in a constructive way. The key concept here is that of inclusive classification or, if one prefers, universality. Naturally, there is nothing the matter with the phrase ‘parish church’ when used of a particular building or group of buildings. It is clear enough what ‘Farnham parish church’ or ‘the parish churches of Exeter’ refers to. Moreover, it is hard to doubt that certain inclusive terms should not be interrogated, or at any rate not too severely, but rather left alone to do a usefully vague job. ‘The Middle Ages’ is one of these terms; ‘culture’ another. However, ‘the parish church’ does not appear to deserve this sort of licence, not least because its historical connotations are too specific. Those who use it intend it to do a specific job.

Before reviewing the matter further, it is worth noting that the present moment is a good one for thinking about how scholars conceptualise the later medieval English parish church. This is because the parish as both cultural and material phenomenon is becoming popular with scholars again after a period of eclipse by the study of monastic and collegiate churches, chantry chapels, and anchorages. Parish studies, and notably their material dimensions, have an impetus not seen for fifteen or twenty years. The current volume is only one manifestation of this. Others include Nigel Saul’s recent book on the relationship between the gentry and the parish church and Gabriel Byng’s study of the parochial economics of building. More popularly, there is John Goodall’s recent collection of mini-essays on individual churches, and Matthew Byrne’s book on parish churches and chapels.⁴ Online, one can turn to the Norwich churches project based at the University of East Anglia and the Corpus of Scottish medieval parish churches based at the University of St. Andrews.⁵ While the latter is not English, it is geographically the nearest thing to a wide-ranging northern study that scholars have (or are likely to get any time soon). This is to name only a few examples: the industry of medieval historians who are sensitive to the value of material culture is adding much more.⁶

In England, one must look back two decades for such collective enthusiasm. This was the first period in which parish studies were professionalised by academic historians and art historians.⁷ Indeed, in 1988, David Palliser could write that little attention had hitherto been paid to the religion of English medieval parishioners, although Colin Platt had produced an accessible and perceptive synthesis on the development of parish churches as early as 1981.⁸ Throughout the 1990s, historians of documents and objects including Eamon Duffy, Katherine French, Norman Pounds, Carol Davidson Cragoe, Beat Kümin, and Robert Whiting prepared or published a large number of books, articles, and theses on, or heavily involving, parish churches in medieval England.⁹ Duffy’s The Stripping of the Altars, published in 1992, was the single most significant stimulus.¹⁰ Duffy was not solely concerned with the parish, but he certainly identified it and its church as the best proving ground for ideas about what he called ‘traditional’ religion. He followed up Stripping with The Voices of Morebath, ostensibly a micro-history, but one intended to reflect both the vigour and locally-rooted nature of late pre-Reformation Catholicism.¹¹ Historiographically, this was an interesting moment: a leading Church historian turning to parishes and their material culture as vehicles of revisionist history where previous scholars—going back almost to the Reformation—had used monasteries and cathedrals.¹² (If Duffy’s choice seems obvious today, given the vast amounts of data available for the parish, it would have been unthinkable before the rise of social history as an academic subject in the first half of the twentieth century.) Yet for all the success of Duffy’s work, academic attention was shifting by the year 2000 towards monasteries and colleges, whose history then looked ripe for revision a quarter-century after David Knowles’s death. Thus it is that Paul Binski’s article of 1999 on the art history of the English parish church, which properly identified its subject as a ‘problem’, has functioned as a sort of place-holder for about twenty years, staking out ground which is only now being substantially occupied.¹³ A volume arising from a Harlaxton symposium on the English medieval parish held in 2002 was effectively an epilogue that proceeded from and summed up twenty years of work.¹⁴ It is possible to view the parochial emphasis of Gothic: Art for England 1400–1547—the major exhibition of late-medieval art held at the V&A in 2003—in the same light.¹⁵

It is reasonable to think that those involved in the current second-wave parochial turn will seek to build on this past work by discovering previously unknown facts and generating new approaches to interpretation. As suggested, this process might healthily include a more searching scrutiny of operating assumptions than past scholars have thought necessary. To substantiate this proposal, it is necessary to say a little more about historiography. While cathedrals and monasteries (particularly abbeys and cathedral priories) attracted scholarly enthusiasm from the sixteenth century onwards, the parish and its church had to wait until the late-nineteenth century, and particularly the early-to-mid twentieth century, to find champions. Studies of parochial buildings and furnishings emerged primarily from antiquarian work on churches in particular regions, and ultimately from the county histories mentioned above. Much was done on furnishings and images using wills and inventories.¹⁶ Two influential publications were Alexander Hamilton Thompson’s companion primers The Ground-Plan of the English Parish Church and The Historical Growth of the English Parish Church published in 1911 and 1913 by Cambridge University Press. Their influence was presumably due to the high reputation of their author, the lucidity of their organisation and argument, and their affordability. The significance of these short, pocket-sized volumes for the subject in hand is hard to exaggerate, although their effects have become submerged in recent publications. Through them, Thompson helped to justify the material aspects of the parish as fit for systematic, rigorous scholarly attention in their own right.¹⁷ He did this in spite of the fact that his personal interests were overwhelmingly documentary as opposed to formal and material. By comparison, the most influential object-focused antiquaries of the time, including William H. St John Hope, produced little influential work on parish churches. Perhaps this is unsurprising in an age so hospitable to the excavation of monasteries and castles.

This is not to say that Thompson’s books emerged from a vacuum. His peers included David Herbert Somerset Cranage, John Charles Cox, Edmund Arnold Greening Lamborn, and Albert Reginald Powys, as well as writers more squarely focused on parochial history.¹⁸ Much of the output was effectively a sort of superior hackwork produced for a lay readership: this was the way, of course, with a great deal of writing on English art and architecture before these subjects were colonised by academics. Without any obvious lull, the parish church baton passed subsequently to the likes of George Henry Cook, Graham Hutton, and Hugh Braun. The publishers entering the field, including Thames & Hudson and Faber & Faber, were such as catered to the self-consciously ‘educated amateur’ with aesthetic leanings.¹⁹ The cloud of authors and books gathered steadily in the wake of the Second World War, peaking in terms of popular influence with Simon Jenkins and his England’s Thousand Best Churches, which, as its introduction explains, includes only buildings ‘in some sense parochial’.²⁰

Publications on parish churches of the last century share three important ideas about their subject. The first is that the parish church constitutes a self-defining category in its own right, whose nature does not need to be investigated. No author ever seems to question what sort of historical or material or spatial sense it makes to consider (say) St. Mary Redcliffe at Bristol (which Elizabeth I is said to have called ‘the fairest, goodliest, and most famous parish church in England’) and Culbone church in Somerset (supposedly England’s smallest parish church) as the same kind of thing. Neither does anyone ask how the medieval parish church located in a surviving monastic nave—as for example at Crowland, Rochester, and Sherborne—fits into the picture.²¹ The assumption of some shared, defining property of ‘parish church-ness’ is expressed in the commonly-stated belief that the real importance of England’s medieval parish churches lies in their collectivity. Hamilton Thompson was frank about this: ‘the parish churches of this country form, as a body, one of the most remarkable historical monuments which any European nation possesses’.²² John Betjeman was too: ‘Their profusion is their greatness’.²³ The matter is just as roundly stated in Simon Jenkins’s and John Goodall’s recent public-facing books. In fact, what is normally under discussion is the south of England and Midlands counties including Leicestershire and Northamptonshire.²⁴ Contrarily, but at the same time intelligibly in light of one of the most familiar clichés of Englishness, it is the idiosyncratic character of the parts which helps to define the whole.²⁵ Some big, some small; some Romanesque, some Gothic; some thatched, others tiled; a stump or a spire; a clerestory or not; here town, there country: each one a loveable eccentric couched in an inimitable setting.

The second idea shared by much of the existing literature is that the parish church is a manifestation of ‘the people’ and ‘the soil’. Predominantly, its patrons and users were ordinary lay-folk, and its materials were local, wrought by local hands and traditional craftsmanship. It is an honest thing, and a frank expression of common and traditional values. Here the medieval life-course played out, from baptism to burial.²⁶ This is not just a popular matter. Paul Crossley has suggested that John Harvey’s great biographical register of English medieval architects is an intellectually-superior product of this impulse, albeit with a reach that transcends the parish.²⁷ As such, the parish church is implicitly and favourably contrasted to the cathedral and the monastery, which (one assumes) are remote, cerebral, and perhaps morally suspect in their size, exclusiveness, regulated customs and other things. ‘Parish/great’ is a core dichotomy for historians of the pre-Reformation Church and its art. Where one is positively demotic, the other is narrowly and latinately grammatical. While the great church smacks of fealty to Rome, the parish church, in its pervasiveness as well as its production and use, expresses healthy nationalism. Thus Greening Lamborn: ‘No Englishman can take a just pride in his race and country who has not learned to appreciate and love them’.²⁸

The third idea, closely related to the second, is that the parish church is somehow—for want of a better word—romantic, or, if one prefers, neo-romantic (romantic with a small ‘r’ anyway). A wider scholarly context for it, which tends to oppose Betjeman and Nikolaus Pevsner, can be found in various recent publications.²⁹ In existing work, this idea usually manifests as a sort of authorial impulse rather than an asservation. Accordingly, writing on the topic is frequently seasoned with poetic sentiment, which ranges from the genuinely gratifying to the painfully innocent.³⁰ The non-architectural sites and broader regionality of churches is often part and parcel of the romance, something the introductions to county surveys of parish churches makes clear. (‘The Churches of Suffolk … stand cresting the gentle hills, or beside the slow moving streams beloved of Constable and Gainsborough’: thus Henry Munro Cautley.)³¹ Here, Englishness per se has a fundamental agency. This romantic impulse is part of a pervasive habit of sentimentalising the parish church, something relatively few authors writing on the churches specifically, as opposed to parish history in general, have been able to resist. With no intention to do so (or indeed awareness of the issue), this brand of sentimentalising nearly always functions as a substitute for querying the choice and parameters of the subject. Jonathan Glanchey, writing on Betjeman, encapsulates the point with an access of optimistic enthusiasm:

These churches, this almost infinite variety of architectural prayer and rejoicing, these stone musical boxes, these national roots, are part of our common heritage, whatever our personal backgrounds and religious beliefs, or lack of them, and remain things of beauty, a joy forever and ever. Amen.³²

Pevsner’s economy and frankness—‘big but also unrewarding’, and the like—can seem genuinely refreshing in light of this sort of thing.³³ Such idealism routinely overrides acknowledgement of the neglectful and destructive treatment medieval parish churches have had at the hands of their custodians. This treatment, too, is familiar territory to the medievalist. In fact, as with the romantic enthusiasms stirred by the churches and their settings, the sentimentalists capture it best. For example, the Victorian rambler Louis Jennings looks back on the restoration of Westham (Kent) as though he had witnessed a man being broken on the wheel, or—a better analogy in view of the building’s size and form—the gutting of a great whale. As he ‘fled in horror from the scene’, he might have been escaping any one of hundreds of later nineteenth-century work-sites. The laconic attitude of the workmen—presumably honest local folk who had grown up in the shadow of the church—struck him as squarely at odds with the barbarity of their task.³⁴

These comments on the drift of medieval parish church studies suggest why any querying of the classification would appear both strange and unnecessary to most people. As a category firmly rooted in both the popular and scholarly imagination, ‘the parish church’ seems just the sort of thing that should not be placed under the microscope, but rather, allowed to exert its influence unmolested by nit-pickers. However, as already noted, it is problematic in ways the self-conscious scholar should not ignore. The core difficulties arise from classifying art and architecture according to an administrative category (the parish). Intrinsically, this is a problem for any marriage of form and function. Pragmatically, however, it hardly matters with reference to (say) cathedrals, or the monasteries or friaries of a given religious order. Although the art and architecture of a religious order or college may bear no inherent denominational stamp (typically it does not), one can normally get a synoptic view of both the historical and material domain when attempting to study it. With this, one can aspire to a coverage that deals evenly with the major aspects of the subject. One cannot do this for the parish church. This is the fundamental problem with the category for scholars. It is worth setting out a few of its practical and theoretical corollaries.

To start with, nobody seems to know how many parish churches the medievalist should count. Estimates of surviving numbers typically range from eight to nine thousand. People were, incidentally, more certain and more wrong about this in the later Middle Ages. Ranulf Higden’s Polychronicon, the most popular universal history in the century and a half before the Reformation, reported 45,002 ecclesiae parochiales in England at the time of the Norman Conquest.³⁵ This number was widely accepted: Edward III’s bureaucrats used it to estimate the sums they could raise from their parish tax of 1371, and it is found copied into monastic manuscripts as one of England’s vital statistics (sometimes with slight variation).³⁶ While this optimistic total was presumably based on some sort of guess—a fact that the suffixal ‘2’ was perhaps intended to disguise—the modern vagueness about numbers stems partially from the fact that it is hard to judge where a building starts and stops being medieval. It may have only a modicum of medieval fabric, and no medieval fittings, left to it. Is an original tower enough to constitute a medieval parish church? While Crondall (Hampshire), which lacks only a medieval tower, is incontestably a medieval parish church, is this true of Dundry (Somerset), whose only old element is its tower?³⁷ Such questions make precise counting hard or impossible.

A simpler cause of ambiguity is the terminus ante quem given to the Middle Ages. In recent studies of the English parish, there has been a tendency to prefer what is in effect a ‘long Middle Ages’, extending into Elizabeth’s reign or beyond.³⁸ (This follows a broader trend in historical scholarship.) An admirable booklet on Norfolk’s country churches, in which the words ‘country’ and ‘parish’ are used interchangeably, counts precisely 659 whole medieval buildings and 245 ruined or dilapidated ones based on a terminus ante of 1700.³⁹ This is to give the period a much longer afterlife than normal. It may seem unnecessary to query such a terminus on the grounds that the vast majority of churches standing in 1700 were built before the Reformation. This would be to suppose that 1700 (or whenever) and 1547 are effectively the same for accounting purposes. However, it is well to remember that many chapels and monastic churches were converted to parochial use at the Reformation or later. Tewkesbury Abbey and Holy Trinity at Hull, the later made parochial in 1661, are examples which suggest how the assumption of a long Middle Ages complicates the task of envisioning the whole.

Here is a conceptually-related problem. Even if he or she could count precisely, the conscientious scholar would still face the problem of deciding whether a church that was reconstructed during the Middle Ages—as they nearly all were, and many more than once—constituted a single medieval building or two (or more) such. It is frequently unclear what church one is talking about before its Decorated or Perpendicular Gothic iteration emerged near the end of the Middle Ages. Little or nothing that would help decide this question can be offered of many pre-Black Death churches. Scholars can mitigate the problem when working on the parish churches of particular periods or regions with a little smart phrasing like ‘the Decorated parish churches of Leicestershire’ or ‘fifteenth-century parish churches in Norfolk’. But this does not solve the essential problem of what materially the medieval parish church can be said to be or have been. People who discuss artistic integration in medieval buildings face essentially the same issue. No single point in the long development of a church interior seems more valid than another. Even if one were to decide that it was, then one would have no more than an imaginary picture of the dominant integrated interior. Willibald Sauerländer has pointed up this question with reference to great churches.⁴⁰ It may be adapted to the problem of interpreting a large body of widely dispersed buildings rather than a single interior. The simple fact is that from a distance, the defining forms and components of these buildings, and thus their identity for art-historical purposes, looks fluid rather than fixed.⁴¹

The definition project also involves a basic choice about whether one counts or excludes ecclesiastical buildings that served communal purposes but lacked independent rectors and baptismal and burial rights. This means, of course, dependent chapels, or ‘chapels of ease’ (as distinct from oratories, which were private). To separate churches and chapels as parochial buildings is not to invent a problem, as most work on parish churches in the later Middle Ages either ignores chapels or pays sufficiently little attention to them to imply authorial belief in the distinction.⁴² Many such buildings were parochial to the extent that they met regular religious needs parishioners had and were served by parish priests.⁴³ Documents often suggest that they were more popular in certain periods than the main parish churches to which they were hierarchically subordinate. According to Hamilton Thompson, the ratio of chapels to parish churches in medieval Leicestershire was one to two, and that county’s parishes are not especially large (a large parish was a normal reason for building a chapel of ease).⁴⁴ Large parishes in northern counties like Cumbria had numerous chapels, whose relationship to the parish church is not always clear: some were evidently quasi-independent.⁴⁵ In some cases, chapels of ease were practically as grand as their mother churches: St. Nicholas’s at King’s Lynn is a familiar example of this, and Holy Trinity at Hull was a chapel of ease to the parish church of Hessle (some two miles to the east). There were also cemetery chapels, located in parish churchyards and often popular with parishioners but structurally independent of parish churches. An example to hand is the (vanished) cemetery chapel of St. Mary at Bures in Suffolk, which apparently contained more than one altar and received copious documented gifts in the thirteenth century.⁴⁶ To exclude such buildings would significantly alter the art and architectural history of the parochial landscape, and seems hardly justifiable on either material or cultural grounds. Yet these buildings cannot be included according to the ‘administrative’ definition of parish church. The term ‘local church’ adopted by historians of the Anglo-Saxon period is one way of getting around the problem, although it would be no less vague for the later Middle Ages than ‘town’, ‘urban’ and ‘country’ church, all of which have some currency.⁴⁷

To study parish churches as if they were a unified corpus involves a practical problem which is hidden in plain sight. This problem is influential and indeed inevitable in many survey books and catalogues dealing with late-medieval art and architecture. Simply put, the vast number of individual churches available for analysis leads to cherry-picking, and thence to questionable inferences (or outright generalisations) about buildings whose structure, embellishment and materials were in fact always different from one another. The most obvious product of this cherry-picking is exceptionalism, or false standardisation, according to which brilliant churches like Long Melford (Suffolk) and Northleach (Gloucestershire) are effectively represented as ‘typical’ of the parish church class.⁴⁸ In fact, as whole structures, Long Melford and Northleach are unusual: a combination of force of habit, regional bias, and old-fashioned aesthetic validity conditions the choice of them and/or others belonging to ‘the best of the bunch’.⁴⁹ Such cherry-picking is perfectly understandable, of course, and is even beneficial if one’s goal is to court public engagement. However, it represents a shying away from the big picture study which the current volume and the conference from which it emerged have identified as a desideratum. Ironically, regional bias also means that exceptional things like the fourteenth-century evidence for internal arrangements at Brigham parish church (Cumbria) and the fifteenth-century Grey tomb at Chillingham (Northumberland) are not factored into even the selective account of the material parish church, let alone broader narratives of later medieval English art.⁵⁰

Turning to theoretical problems involves some blunt spelling out, and reiteration of things already said. Fundamentally, the parish church category seems resistant to theoretical speculation. At least, it is challenging to think of a current of cultural theory hospitable to it. Theories of materiality, aesthetics, economics and so forth naturally cut across such administrative categories. No traditional art historical theory is neatly compatible with it. Approaches through form, iconography and style are blind to administrative categories like ‘parish’, ‘cathedral’, ‘monastery’, and ‘friary’. So, by and large, are ideas about human agency, including those that involve making, patronage, and interpretation. Style criticism alone will not show whether a given tracery pattern or moulding profile is in one administrative category of building or another. A knowledge of iconography does not help to situate a sculpture or painting of obscure origin. It has never been argued that the intention and reception of an image included a distinctively parochial quality. While everyone realises this, some readers might think that a broadly-based study of the parish church can sidestep such approaches on the assumption that art history is rapidly collapsing into a form of cultural inquiry which can do without technical knowledge. However, at the risk of stating the obvious, the parish church is no more compatible with the currents of anthropology, sociology, linguistics and so on which feed cultural history. No more compatible, and evidently of no greater interest, for very few parish churches are noticed at the cutting edge of recent and current scholarship. The reason for this is probably that the domain is thought to be intellectually flat, in part (at least) because it is ‘local’.⁵¹ It is certainly the case that much that is captivating about late-medieval parish churches—the porches of East Anglia, the towers of Somerset, the spires of the East Midlands—is also philosophically unstimulating. (Beauty is affective but not necessarily intellective.) If this claim seems unjustified, then it will nevertheless be admitted that such architecture has inspired extremely little theoretical speculation to date.

This implies that ‘parish church’ is a term of convenience which encourages, or anyway facilitates, intellectual complacency by masquerading as a respectable premise from which to advance. Once chosen, scholars can get on with studying the few outstanding or well-documented examples that motivate them or support their case without considering the validity of the parameters they have set themselves. Like it or not, this is the attitude reflected in the introductions to almost all general books on parish churches. It leads immediately to a matter that is basic and important to how art historians work. This is simply that, however materially and geographically broad it seems, the concept ‘parish church’ is too exclusive for scholarly ambitions. If scholars could reconcile themselves to the indistinctness already discussed, and confine themselves to the parish domain, they would miss the chance to contribute broadly to art and architectural history. Not only might they do odd things like leaving out St. Nicholas’s at King’s Lynn, they would also disqualify any relationships between art and architecture that happens to be parochial and non-parochial. Even if they took account of such relationships, they would help to perpetuate a category they could not clearly define: a ‘soft’, romantic, flabby category, speciously justified by medieval legal terminology but intractable both in practice or theory. Moreover—in effect if not intention—a chauvinistic category, which tacitly rejects the possibility that other European countries could show anything like the collective glory of England. There are, in any case, very few transnational studies of the material parish church.⁵²

This is supposed to read as a provocation, not so much in order to repudiate a calumny as to recognise soft spots in a working assumption that is both common and deep-rooted. It is no evasion to conclude by insisting that this is all the current essay has been intended to do. Indeed, it is easy to admit that the idea of a unified domain has a dignity against which both practical and theoretical objections seem irrelevant. Before all else, this idea embodies an even respect for all of the parts that denies judgements about their aesthetic or historical worth relative to one another. That respect is manifested in the claim of Peter Lasko and Richard Fawcett that ‘the loss of any part … would have inevitable repercussions for the whole of English architectural history, and disastrously diminish the value of the evidence of surviving structures’.⁵³ Theirs is simply a job-conditioned way of saying that individual parish churches are elements of a big, incalculably-valuable thing which is compounded of history, materials, and affects, is intensely meaningful to many people, and will go on being so. To crib from John Steinbeck, ‘none of it is important or all of it is’.⁵⁴ There can only be one response to that proposal.

It need hardly be added that the category ‘parish church’ is useful where adopted with open eyes and a flexible approach to shape. This is indeed the way it is often approached. For example, to study the parish churches of a single town or city that had a lot of them—Lincoln or London, Norwich or York—surely promises the art and architectural historian cogent and useful results, particularly if they are regarded as components of a broader ecclesiastical environment that included religious houses, chapels, anchor-holds and other things. Sampling across a broad range can help clarify larger matters of architectural history. Thus, for instance, by demonstrating a radical decline of stone vaulting in parish churches from the later thirteenth century, Larry Hoey helped to confirm a general hunch about the priority that builders in the Decorated and Perpendicular periods gave to window tracery.⁵⁵ That is, once complex tracery was invented, the money available in the parish for making a splash went to windows, which could be seen from outside as well as within, rather than vaults. This seems a thoroughly useful conclusion, generated by studying a small sample of parish churches. These are only two of many ways in which the virtue of studying samples made coherent by the label ‘parish church’ could be expressed.

On the whole, however, art and architectural historians will better maintain their distinctiveness as a scholarly cadre, and thus serve their compromised, unstable discipline, by studying phenomena which transcend socio-historical categories like ‘parish church’. It can, at least, do no harm to think about what such categories could mean in the context of a distinctively art-historical discourse.