More is known about how things were cooked and served than is known about the details of cuisine, although more could undoubtedly be gleaned from a detailed study. A variety of means of preparing food is represented in the Iron Age record. Hearths are present in many domestic structures. As well as providing heat and light, they would have been the principal cooking and smoking resource; a hearth allows a variety of different cooking options, from bubbling cauldrons or spits over the flame or cooking pots set on the hearth to grilling on the stones or baking in the ashes. Rare oven-like structures have also been identified such as the one at Old Scatness (Dockrill et al. forthcoming); a study of hearths and ovens is currently being undertaken by J.R. Summers (in prep) as part of his PhD research. An alternative to hearths is cooking pits (e.g. Rideout 1995), and in some areas and periods they are locally abundant. This variation between the hearth and the pit as the dominant means of cooking merits a detailed study, as it represents very different traditions of cooking and thus styles of cuisine. While burnt mounds are primarily a Bronze Age habit, the presence of fire-cracked stone in abundance on Iron Age sites indicates that hot stone cooking technology remained a key feature (Barber 1990), but now on-site rather than off-site as in the Bronze Age. The presence of stone-lined tanks or cists at several sites has been interpreted as food (or water) storage or processing features (MacGregor 1974; Dickson and Dickson 1984; Dickson 1994). The whole area of cooking apparatus merits detailed research on a regional or national level.
Reconstruction of diet, and the balance between meat and plant resources, is very difficult. More direct evidence comes from rare evidence of coprolites. Some of the coprolites found at Howe were thought to be of human origin but it was not possible to analyse their contents (Ballin Smith 1994). A collection of human coprolites from the well at Warebeth broch (Dickson 1989) gave surprising results as they contained a large proportion of meat, plant remains being much scarcer than expected. As a result they were interpreted as being the result of atypical meals, possibly from a time of cereal shortage, and were thought unlikely to be representative of the everyday diet of the inhabitants of the site (although diet probably varied seasonally). Another direct (and relatively new) method of determining aspects of diet is the examination of starch granules trapped in dental calculus. In a pilot study Hardy et al. (2009) identified cereal starch in dental calculus of skeletons from Pictish Tarbat.
Stable isotope studies are also useful dietary indicators. At early medieval Newark Bay in Orkney, Richards et al. (2006) found there was higher than expected marine input into the diet and considerable variation between individuals. In contrast, in Iron Age East Lothian, Jay (2005, 242-243; Jay & Richards 2007) found slight variation between sites but a great deal of homogeneity at any one site and no indication of a significant proportion of marine input into the diet. There was also little indication of diversity between individuals in terms of diet, but the topic is worth pursuing as more skeletal data becomes available.
Such studies give an idea of the ratio of different elements in the diet, but say little about cuisine: there are many ways to cook a pig. In favourable circumstances, the animal bone assemblage can give some idea of how meals were prepared - for instance, whether bones were charred, or affected by boiling. Likewise butchery styles and bone representation can give some idea of the cuts of meat used; for instance, at the Dürrnberg (Austria), the evidence suggested stews were the main dish, as these extract maximum value from poor meat (Stöllner 2003, esp. 170).
Artefactual evidence, specifically pottery, ought to give insights into the preparation, storage and consumption of food, though in Scotland it is rarely interpreted in such terms. The tendency has been to look on pottery primarily as a dating tool, at least in the Atlantic regions where it is commonplace, or to bemoan its absence in other areas. Yet these two very different situations of pottery use also ought to inform us about people's cooking and eating habits. While pottery sequences have been studied in terms of typological development (eg Young 1966; Campbell 2002), there has been less concern over functional development and diversification. A valuable piece of research would be an overview of a region's ceramics (and indeed other vessels) in terms of vessel capacity and form, the latter interpreted in terms of likely function. Work in southern England on vessel size, form, and the analysis of context groups in terms of function, provides models for such studies (Woodward 1997; Hill 2002; Morris 2002), although the generally fragmentary nature of Scottish assemblages does limit the potential somewhat (A MacSween, pers comm.) Campbell (1991) provides a rare attempt to interpret potential meanings behind the decoration of pottery (see more generally Woodward 2002).
The other area of investigation is the rarity of pottery in much of the lowlands. As Willis (1999, 83-90) has shown in the analogous case of northern England, the area was not aceramic but pottery was uncommon. The implication must be that vessels in a range of other media were used. Of these, there are rare survivals of wooden vessels in a diversity of forms, from cups to massive buckets, and suggestions of other materials such as birch bark containers which could readily fulfil the consumer end of the spectrum (Earwood 1993). The form of some vessels resembles bread troughs in more recent times, for kneading dough, though while a number were found to contain have actually been recovered with "bog butter", a generic term for a range of animal products, some dairy, and some representing products such as tallow (Hunter 1997, 128-9; Berstan et al 2004). This serves as a valuable reminder of the range of secondary products available from animals - milk, butter, cream, blood, sinew, gut and so forth.
Selection of Iron Age pottery from Orkney, © NMS
Pottery and wood were not the only options for vessels, but evidence for other materials has not been fully studied. The occurrence of vessels in other materials is less certain. Leather vessels are likely, though near-impossible to prove; wicker vessels are known (eg from Howe; Dickson 1994, illus 82), but their form constrains their use. Stone "cups" from northern Scotland are more likely lamps (Stevenson 1966, 28), but various forms of stone bowl are known, including rare steatite examples in the Northern Isles (Forster 2009). Copper-alloy vessels were also used, from cups to cauldrons (MacGregor 1976, 147-152), although their while the original incidence of copper alloy vessels such as tankards or original prevalence cauldrons is hard to assess. Occasional pottery skeuomorphs of bronze vessels indicate the latter's perceived value: examples include the burnished carinated early Iron Age pottery known from Clickhimin, Shetland or a small cup with bosses imitating rivets from the Howe (Hamilton 1968, fig 19.1-5; Ballin Smith 1994, 248).
The social context of consumption is also elusive and tends to be dominated by exceptional circumstances. At some sites deposits of animal bone material have been interpreted as having a special function i.e. they are not part of the everyday domestic refuse but may have been involved in some form of special event. These may be sacrificial or ritual deposits or the remains of feasting. A series of examples may illustrate the phenomenon. High Pasture Cave had an unusual species distribution compared to many sites and a number of examples of articulated animal remains (Drew 2006). An array of ritual pits containing articulated and cremated animal remains wasere excavated at Sollas in North Uist (Campbell 1991), while Davis (2000) identified a deposit made up mostly of calf hind leg elements in close association with human cist burials at An Corran, Boreray. and Cussans and Bond (forthcoming) interpreted a deposit of prime meat age cattle and prime joints of pork and lamb making up the primary ditch deposit at *** as the remains of a community feasting event. A perceived significance to animals is seen also in other phenomena, such as the use of cattle metapodia to define a hearth at Bornais, S Uist, or deer jawbones at A'Cheardach Bheag (Mulville et al. 2003; Mulville & Thoms 2005, 241).
There has been little detailed interpretation of the meaning of such events - contrast the innovative work of Jones (2007) on social interpretations of the pit deposits at Danebury, or Hill (1995) on deposits in Wessex, which offer exciting prospects of what could be obtained from such data. Campbell's (2000) structuralist interpretation of the Sollas deposits, drawing together a range of sources of evidence, is an enlightened example of recognising the patterns which lurk in such data, offering models for debate about meaning.
An area which has seen little detailed study is the question of alcohol, although the evidence would only survive in very favourable circumstances, such as burials or hoards where vessels could be sampled. It is likely that beer was a staple drink, and possible also that fruit wines were made although hard evidence is elusive.
There is a poor understanding of the nature of cooking practices and cuisine in the Iron Age, although much raw evidence exists. Specific examples include: patterns in preference for hearths rather than pits as cooking devices; interrogation of animal remains for information on cooking practices (butchery, charring etc); size, form and thus function of pottery and other vessels
A range of scientific approaches have considerable potential for the study of food, cooking practices and cuisine: organic residue analysis of pottery; isotopic study of human bones in terms of diet and the balance of arable and pastoral resources, and animal bones in terms of mobility; coprolites
Patterns in the deposition of animal bone can provide important information on social habits such as feasting and insights into beliefs, but this remains an area ripe for detailed research, with only a few pioneering studies so far.