An article in VA 28 interpreted several detached buildings dating to the fifteenth or early sixteenth centuries as kitchens rather than the independent houses such structures have commonly been assumed to be. What follows is a challenge to the new interpretation.
The recent important article by David and Barbara Martin claiming that certain buildings in Sussex are detached kitchens (Footnote 1), not the houses they might be taken to be, raises questions that need fuller discussion (Footnote 2).
First, how is a kitchen defined and by what criteria of plan or structure can it be recognised? For a definition, 'the room or part of a house where food is prepared and cooked (Footnote 3)' would find universal acceptance. The corresponding architectural criteria are more complicated. To confine the argument for the moment to detached buildings, the big square kitchens of monasteries and great houses with louvred lantern and multiple ovens are easily recognised for what they are. They have provided the model for the interpretation of much smaller square buildings standing close to manor houses or farmhouses for which any function other than a kitchen can scarcely be envisaged. A medieval example was recorded by Hewett at Little Braxted Hall, Essex; it was some 6.7m (22ft) square and 4.6m (15ft) high to the eaves (Footnote 4). Such a structure cannot be a house even though the cook or a scullion almost certainly slept there, in the way that people seem generally to have lived and slept where they worked in the middle ages and much later (Footnote 5); and any lack of an oven or ovens is offset by the evidence of a hearth. To go back to the definition, a complication arises from the inclusion of food preparation, as the name 'back kitchen' applied to an unheated room used for this purpose in some Welsh farmhouses shows. A second room, unheated, may therefore occur in some circumstances, although what fittings might be expected to identify it or in what relation it should stand to the cooking kitchen are hard to ascertain. Two-room structures are more problematic than simple square ones. In the light of the discovery by Fox and Raglan of a few detached kitchens in Monmouthshire others were claimed by the Royal Commission working in Dorset. One such, at Corfe Castle, was a little building of one storey and attics comprising two rooms each about 3m (10ft) square, one unheated – storeroom, cook's room, preparation kitchen? – and the other with an internal lateral chimney stack that reduced the width to some 2.6m (8ft 6in), thereby producing for that room the proportions of some medieval kitchens, which are wide but comparatively shallow in front of the fireplace. 'John of Gaunt's kitchen' at Canford Manor (Dorset) is a large example and many much smaller late sixteenth and seventeenth century kitchens have similar proportions (Footnote 6).
But the instances brought forward by the Martins are more dubious, as their comment about 'the house-like appearance of detached kitchens' implies. So what criteria do they use to distinguish between the two kinds of building? Setting aside the single square example, the others (again) 'closely resemble small houses' so that 'it is often only the location of these structures close to the rear of a main house' that reveals their function; and in some examples 'the internal arrangement of the kitchen indicates very clearly its subservience to the main dwelling' (Footnote 7). Now for the first of these criteria to carry much weight it is necessary either to establish that two houses in one yard – such as might indicate joint occupancy of a farm (Footnote 8) – did not, in other instances, stand in much the same relation to one another as kitchen to house, or to demonstrate that the unit system, as it is called, is an illusion. Neither kind of demonstration was attempted. Similarly with the internal arrangements of these supposed kitchens: only if they differ in some distinctive way from those of unit system houses does their evidence count for much.
The latter point can be dealt with briefly because so little relevant evidence is offered. Two of the alleged kitchens, 12-13 High Street, Battle, and Beestons, Warbleton (Footnote 9), appear not to differ from an orthodox house in any significant way, if at all. In the case of Comphurst, Wartling, the long walls of the 'kitchen', aligned on the end partitions of the hall of the adjoining medieval Wealden house, carry a social meaning that is hard to define exactly but is certainly not a message of subservience. Nor is a subordinate relation obvious in the lack of alignment between the axial passage (between two service rooms) of the 'kitchen' and the cross passage of the 'house'. Just such an alignment appears at Darwell Beech, Mountfield, where a passage set at the end of the 'kitchen' and to one side of it faces the cross passage of the 'house', and here the size of the fireplace and the proportions of the room it serves of are indeed appropriate to the suggested function; although there is a contrary reason (below).
Before looking further at the way two buildings, both having the appearance of a house, might be set out in relation to one another, a general point deserves consideration: for what kind of household did a detached kitchen provide, both in the main house and under its own roof? To take first the main house – the house, in the Martins' terms – it seems, in the two examples where a plan is provided, to be remarkably close in size to the kitchen. Comphurst, Wartling, is a Wealden, unfortunately of uncertain length and width because the plan as presented is incomplete, but two criteria provide indications of the relative importance of the two structures. One, a very important aspect of medieval domestic buildings, is width; and at about 4.8m the kitchen would qualify as a small house, even in the generous terms to be expected of its kind in Sussex. A second and more significant criterion is the presence, on two sides, of a jetty; why should a building of menial function be provided with so conspicuous a display feature? The latter point in particular swings the interpretation in favour of a house and would, in a free-standing farmhouse, emphasise that the household was of some standing; why not in this case? Darwell Beech, Mountfield, the second example, permits a very approximate comparison of area for the two buildings, excluding upper floors for which the information is inadequate; the house is about 75sq.m, the kitchen 59sq.m, giving a ratio of about 5:4. Even more remarkable is the comparative width of the two buildings, the alleged kitchen being wider than the house it served, yet width is a good indicator of relative importance in vernacular houses. And Darwell Beech is matched by a house at Ashley, Hants, which has a ground-floor ratio to its supposed kitchen of about 9:7 (Footnote 10), and at a guess the relative sizes of the two buildings at Comphurst would not be very different. This nearness in size prompts questions. Why should a farming household which, it seems to have been assumed without actually saying so, was a nuclear or at most a three-generation family, require a kitchen so large relative to the size of their own house? And was it difference in household size that made small square kitchens adequate for some households but necessitated house-like kitchens for others?
Documentary evidence is brought to bear on these matters (Footnote 11). A Wealden house at Great Worge, Brightling had, in 1567, a building described as a kitchen. It had a total of three rooms, one to dress meat in, which is the kitchen proper, and a bakehouse and milkhouse, the last two being 'loathed over', that is, they had a first-floor loft or chamber. But which? It is important to know what the term 'loathed over' really means – it sounds akin to lofted over but requires elucidation – because the kitchens for which drawings are provided have upper storeys, whereas the uses of the 1567 kitchen suggest lofts and there is no hint of rooms used solely as accommodation. An Essex reference of some two hundred years earlier describes a manor house comprising a hall with an upper chamber at one end and a single-storeyed service end – a fairly typical smallish open-hall house – as having a kitchen with a guest chamber at the end and a bake-house. Since the manor belonged to Westminster abbey the guest chamber was probably for visiting officials on visits of inspection. For Sussex houses generally more regular domestic uses are suggested, either lodgings for servants or dower quarters, as well as general storage. None of this goes to the heart of the problem. General storage is a pointless expense in a jettied building unless, standing in some prominent public place like a market (Footnote 12), it makes a statement about its owner. Lodgings for servants are equivalent, in the examples illustrated, to servants' halls in the sense the term is used in country houses, and amount to separate households. Dower quarters, which could probably be exemplified from wills and probate inventories, seem often to have amounted, in the seventeenth century, to a room with access to other facilities and, given the uncertain duration of the life span, to have been often a makeshift arrangement rather than the separate house built by greater landowners for that purpose. Very fairly, the Martins illustrate their own difficulties of interpretation with the 'one of the larger surviving detached kitchens' at Beestons, Warbleton, which is described in documents of the 1630s variously as a kitchen or house and of which they say that use as a house would explain its survival, 'for it was not converted into an oasthouse until the eighteenth or nineteenth century'.
That remark reveals one major weakness of the Martin’s article that is an inevitable consequence of their admirable determination to publish in the Rape of Hastings series large amounts of material in a small compass and at very modest price. This aim necessitates showing the buildings in diagrammatic form without any of the major alterations that are usually needed if they are to be kept in use and so enabled to survive, yet understanding of a house is enhanced by knowledge of its later development; rooms often preserve comparable functions despite such drastic changes as the insertion of a chimney stack or the replacement of one kind of stack by another (Footnote 13). These can provide clues to the earlier uses of rooms or parts of them that the authors may have noticed but, if they are mentioned at all, have to be taken on trust by the reader. In the case of Beestons, what is the architectural evidence for its continuance as a house and for the date of its conversion to an oasthouse? And at Darwell Beech, how did the two buildings develop before c.1730 when, we are told, they were first joined? Another problem raised by the Martins' paper is implied in the illustrations of Combe Manor, Wadhurst. This house was built with an attached but not, apparently, intercommunicating kitchen which is no more than 18sq.m in area, yet the purely domestic area of the house is much the same as Darwell Beech. So why did the former need only a small kitchen and the latter require one more than three times as big? The answer to that question cannot relate to the preparation and cooking of food or to the storage of the utensils needed for those processes unless it can be shown that the respective households were of markedly different size. That leaves the existence of a second household as the most likely explanation, with the consequent need to speak of a second house, not simply a kitchen.
The questions prompted by these Sussex houses recur with the less well- preserved evidence assembled by a Royal Commission team in Kent (Footnote 14). Unfortunately, at the few relevant places the alleged kitchen has been gutted or the coeval house has disappeared, so that only two comparisons of ground-floor area can be made. One, the fourteenth-century stone-built Yaldham manor house, Wrotham, has next to it a timber- framed building about 13m x 6.5m whose status as a kitchen is thought uncertain in both the Historical Analysis volume and the Gazetteer on the grounds of its size. The other instance, now separate houses called Old Gilwyns and Gilwyns, Chiddingstone, have proportionate ground-floor areas of about 9:5 – a somewhat more credible relation between them but still way above the proportion of any house to a square kitchen. Nevertheless, the simplicity of the smaller building – two smoke-blackened bays with no sign of other accommodation (Footnote 15) – is consistent with a simple functional explanation. A third example, a two-storey structure which 'may have been 'an unusually elaborate kitchen' is at Mead Manor, Sturry where the manor house itself has disappeared (Footnote 16); with a dendro date of 1475 'it is too early for a 2-storey dwelling', hence the conclusion that it was 'an ancillary range, possibly a kitchen'. Why a kitchen, a type hitherto identified as a square building open to the roof, should develop into a two-storey structure floored throughout, at a time when houses were still built with open halls, is not explained. If the alternative interpretation of such buildings as second houses is considered, it is clear that the status of the putative second or minor household will have varied in relation to the major one. It should be noted that where two households existed at one farm their houses, and by inference they themselves, are hardly ever exactly equal; Arnford, Long Preston and Askrigg Hall (Yorks) are the only instances known to me (Footnote 17). Where a secondary building clearly incorporates a kitchen on the evidence of the form of fireplace or of fittings but includes accommodation over and above what is directly explicable by that function (cook's room, spence), especially upper chambers, or has a finish, such as a jetty, superior to what a menial function requires, the possibility of its having accommodated a second household of slightly (or considerably) lower status than the main one can be considered. In such a case the fireplaces of the main house require examination to see if they were ever of a form suitable for cooking or have been rebuilt from such a form.
A further general point concerns the placing of two houses in relation to one another. The number of ways in which two rectangular structures can be set out to express the social relations between them is considerable and every one carries a shade of meaning. Two equal houses facing the same way may form a single range (Arnford); they may face the same way but one is set back so that they join only at the corner (Breconshire, Hertfordshire) (Footnote 18); they may be of different size and set at right-angles, either abutting or separate (Monmouthshire) (Footnote 19); or be separate, either parallel or set end to end at a slight angle to one another (Kent – Yaldham, Wrotham, perhaps?); and in any variation of these settings-out and with or without intercommunication. And if the nuances of social relations are to be properly inferred the siting of the individual buildings must be related to the approaches to them and to the farm at the time they were built or, in default of such evidence, by assessing the possibilities from the modern topography. The architectural arguments have to be weighed against the documentary evidence mentioned earlier and more to the same effect in Kent, e.g., 'my kitchen with a solarium' (Cranbrook, 1490) and 'my new kitchen...with the chamber over the said kitchen and parlour' (Tenterden, 1512) (Footnote 20). Both these instances provided accommodation for widows, a possibility mentioned by the Martins (Footnote 21). That purpose-built provision for a situation that by definition was comparatively short-term should have been made at vernacular level seems unlikely, bearing in mind the not uncommon mention in wills of a chamber for a widow with access to the hall, or similar phrasing; rather, that particular situation may have found a solution in a readily adaptable building erected for a different purpose.
A more telling argument is founded on the lack in wills or leases of any hint of two houses or households. This problem recurs with gentry seats which have or had two houses, for example Temple Dinsley, Preston, Herts (Footnote 22), and even in Wales, so rich in genealogical information and well documented in terms of wills and leases, little enough has come to light. Only architectural investigation revealed that two quite distinct gentry houses, Edwinsford Uchaf and Edwinsford Isaf (Caerns), abutted one another and had been concealed within a single Victorian mansion (Footnote 23). Similarly with farms where two more or less equal houses are found or where a single elongated house has two symmetrically placed entrances (Footnote 24), evidence of joint occupation has not been found; yet they exist and are hard to explain in any other terms than the unit system. None of what has been said is to deny categorically that buildings of domestic appearance, whether in Kent, Sussex or elsewhere, may have functioned as kitchens for a second and more important household, only to say that any such conclusion needs much more detailed argument to substantiate it. Perhaps we should envisage second houses in a considerable range of size, from the single room, as with some of the Monmouthshire examples, to an imposing five-bay early Georgian house abutting the rear corner of an earlier one of much the same size (Footnote 25), and allow that some of the smaller ones may have served as kitchens to the principal house without being purely menial in character like the square detached kitchens of the middle ages. But if the latter possibility be accepted, its implications for family structure ought not to be ignored.