Physical appearance

The Bronze Age field ditches at these south-east Suffolk sites were morphologically similar. They were generally narrow and shallow, ranging from 0.3–1.9 m wide and c. 0.1–0.7 m deep but, much more typically, about 0.5–1.0 m wide by no more than 0.2–0.3 m in depth. Profiles were usually moderately sloping and rounded (Fig. 8), though the (also more substantial) main trackway/enclosure ditches at Walton Green were more ‘v’-shaped. Notwithstanding their overall rectilinear layout, individual ditches often exhibited a slightly sinuous/meandering appearance.

Fig. 8. Example field system/trackway ditches: 1. Trimley trackway ditch fully excavated for finds retrieval, view south-east (photo: Clare Jackson); 2. Martlesham Ditch 27, view north (photo: Tom Woolhouse); 3. Martlesham Enclosure 1 double-ditched boundary with entrance to right of photo, view north-east (2 m scale) (photo: Tom Woolhouse); 4. Martlesham Ditches (from left to right) 20, 17, and 19, view north-east (1 m scale) (photo: Jonathan House)

Upper portions of some ditches at Brantham had probably been lost to plough damage, accounting for much of the intermittent appearance of the shallower boundaries, particularly along the field block’s south side. Discontinuous ditches at Felixstowe and Trimley are most likely a result of their being cut though the loess-derived subsoil at those sites but not always seen at that level during machining. However, it is worth noting that at some Thames Valley and Cambridgeshire sites there are indications that boundaries were originally constructed from a series of short, interrupted ditches, like these, that were only later made continuous (Yates Reference Yates1999, 165; Leivers Reference Leivers, Lewis, Leivers, Brown, Smith, Cramp, Mepham and Phillpotts2010, 209; Phillips & Mortimer Reference Phillips and Mortimer2012, 20). It is also worth bearing in mind that if, as seems likely, it was a bank/hedge that was important in demarcating the plot of agricultural land (see below), then it would not matter if ditches were discontinuous – their function was primarily as quarries to make the bank (Pryor Reference Pryor1998, 145; Lambrick Reference Lambrick, Lambrick and Robinson2009, 60). On the other hand, some other gaps in ditches were certainly ‘real’ entrances: Enclosure 1 at Martlesham had a 2 m wide entrance opening onto a trackway (Fig. 8.3), the corner entrances to the Brantham fields, including the 12.3 m opening onto the western trackway, appear genuine, there were entrances 6–7 m wide into a couple of the fields/enclosures at Walton Green (FEX 312, Carvey Reference Carvey2018, 17–18) and there was a c. 6 m entrance between the Leiston trackway and fields on its west side.

Across all the sites it was rare for there to be much discernible variation in ditch fills, most exhibiting gradual infilling through natural silting and weathering, leaving visually homogeneous deposits, a process no doubt exacerbated by subsequent mixing from rooting and other soil processes to which the generally shallow (and not deeply buried) features were exposed. In landscapes where holes in the ground fill with slumped/windblown sand and silt in a matter of weeks, it is difficult to see how systems of shallow field ditches could have remained visible/functional unless they were regularly cleaned out or recut, or if they were accompanied by banks or hedges that were longer lasting landscape features and which might have helped counteract soil erosion.

There was some direct evidence for recutting and redefinition of ditches at Martlesham, where an earlier demarcation of the field boundary survived in places beneath/alongside the northern part of Ditch 31 (Ditch 34). A later ditch (Ditch 30) on the same alignment as Ditches 31, 32, and 33 in turn cut Ditch 31. At the south-western terminus of a very slight ditch (Ditch 20) were traces of two equally slight, earlier ditches on the same orientation (Ditches 17 and 19), the continuations of which had been entirely removed (Fig. 8.4). There is also some circumstantial evidence for hedges, for instance, the double sets of ditches surrounding some of the enclosures at Martlesham (particularly Enclosure 1 but also 5 and 6), which are too closely spaced (c. 1.5–2 m) – especially when loss of the contemporary topsoil and upper parts of the ditches is factored in – to have formed trackways and are therefore more likely to be ditches either side of embanked hedgerows (see Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009a, 45; Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009b, 245; Lambrick Reference Lambrick, Lambrick and Robinson2009, 58). Alternatively, banks could be topped with fences/stockades to keep livestock in and predators out (Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009b, 249). Similar narrow-spaced double-ditch arrangements surrounded some of the (larger) rectangular enclosures in the Bronze Age field system at Fengate (ibid.); Francis Pryor interpreted these as ‘internal droves’ (Reference Pryor1980, 23) and later – with some supporting phosphate evidence – as sheep drafting races (Reference Pryor1996; Reference Pryor1998, 100–5; Reference Pryor2001, 409–10, 420) but this generally seems unlikely in view of their narrow widths and positions surrounding three or more sides of a given enclosure (Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009b, 245).

Martlesham field Ditch 30 had a pronounced and otherwise inexplicable kink at its north-east terminus, perhaps where it avoided a tree, as at Bradley Fen, Whittlesey, Cambridgeshire, where the wetland edge boundary ditch avoided a tree whose stump was preserved in the adjacent bank (Knight & Brudenell Reference Knight and Brudenell2020, 151–2, fig. 4.10); trackways at Heathrow Terminal 5 similarly zigzagged around trees (Leivers Reference Leivers, Lewis, Leivers, Brown, Smith, Cramp, Mepham and Phillpotts2010, 142). A natural-looking linear feature [1052]/[1054] that extended intermittently alongside Martlesham Ditch 46 could be a remnant of a hedge (Figs S14–S15). The close spacing (c. 1 m) of Ditches 5/6 and 7 at Easton (Fig. S17) might also indicate an embanked hedgerow.

Indeed if (parts of) these field systems were intended to contain livestock (see below) then hedges must surely have been integral components: even allowing for a missing depth of Bronze Age topsoil, the generally ephemeral ditches and small banks that the arisings from them could have formed would not present a significant barrier to animals. In many cases the principal purpose of the ditch may have been to generate a low bank of loose soil into which cuttings for a hedge could be planted, either for stock control, as a clear demarcation of land allotment/tenure, or both (Pryor Reference Pryor1998, 85–7; Yates Reference Yates2007, 121; Lambrick Reference Lambrick, Lambrick and Robinson2009, 58; Leivers Reference Leivers, Lewis, Leivers, Brown, Smith, Cramp, Mepham and Phillpotts2010, 142). If the main boundary feature was the hedge sitting atop low upcast bank, this could help to explain the generally limited evidence for ditch maintenance/recutting – the ditch’s job was done. This interpretation of the ditches as components of barriers, whether practical, visual, perceptual, or all three, can be weighed against the alternative: seeing the ditches primarily as field drains. That is, overall, difficult to square with the context of the ditch systems on mainly free-draining sandy geologies, the general lack of slope at most of the sites (based on both surveyed levels and available LiDAR imagery; DEFRA 2019), and the aforementioned tendency for shallow features on these soils to rapidly infill, especially if/when field surfaces are bare. Cutting through the more water retentive loess subsoils of the Felixstowe peninsula to reach the underlying sands and gravels may have been of some help to drainage (see Yates Reference Yates2007, 138) but this need not have been the ditches’ sole or primary purpose.

In the main, sizes of individual fields cannot be reconstructed owing to piecemeal exposure. Whole fields/enclosures were only seen at Martlesham and, perhaps, Ipswich Academy. Minimum sizes for the larger land units can be roughly calculated from what was revealed within the excavation areas at Brantham (Fig. 2: west field: 2500–2800 m²; central: 3500 m²; east field: 2300 m²), Trimley St Mary (Fig. 2: Enclosure 1: 3200 m²; Enclosure 2: 3400 m²), Felixstowe Academy (Fig. 2 & S3: 7200 m² between Boundaries 5 and 6), Martlesham (Fig. 5: 9400 m² between Ditches 30, 31/35/36, and 37), Easton (Fig. 5 & S11: 2500 m²), and Leiston (Fig. 7: 2375–just over 3000 m²). If the two opposing sets of ditch alignments at Ipswich Academy are interpreted as contemporary, then square/rectangular fields of c. 3250 m² (Fig. S6: between Ditches 1, 2, 6, and 13) and approximately the same area for the block between Ditches 1, 6, 13, and 16, can be identified, with more closely spaced ditches demarcating smaller enclosures near to the identified structures. Fields in these systems therefore varied from under an acre (<0.4 ha) to upwards of 2 acres (>0.8 ha). Several of the more complete prehistoric fields at Flixton appear larger: in excess of 5 acres/2 ha for the plots between Ditches 4, 6–8, and 11, and 14–16, though evidence from adjoining unpublished area FLN 091 suggests that ditches sub-dividing the latter field, at least, may have been lost (compare Boulter Reference Boulter2022, 4 fig. 1.3 and 64 fig. 3.45 with Boulter Reference Boulter2017, 40 fig. 11; Fig. 7). Nevertheless, the overall impression is of quite small fields, which may have important implications for the scale and social context of agrarian production (see below).

These small plots compare closely with the ‘Celtic fields’ of the southern chalklands: on Salisbury Plain some fields are only a few tens or hundreds of square metres and rarely exceed 0.5 ha (McOmish et al. Reference McOmish, Field and Brown2002, 56). At Bradley Fen, those Middle Bronze Age field plots that could be completely reconstructed ranged between 0.12 ha and 1.04 ha, though some incomplete plots measured at least 1.33 ha. Strip-like arrangements of fields, like those seen at Brantham (c. 40 m wide by 80 m+ long) and Leiston (just under 30 m wide by 100 m+), are also a common component of many Bronze Age sub-divided landscapes. Regular field strips at Bradley Fen were 30–60 m wide and the largest field was up to 160 m long (Knight & Brudenell Reference Knight and Brudenell2020, 147–8, table 4.1, figs 4.5–7). Strip-fields were also part of the systems at Barleycroft Farm/Over (there, c. 50–80 m wide: Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009a, 53–5, figs 2.19–20) and Clay Farm, Trumpington (60–100 m: Phillips & Mortimer Reference Phillips and Mortimer2012, 19, fig. 6), both in Cambridgeshire, South Hornchurch, Essex (45–50 m: Guttman & Last Reference Guttman and Last2000, 326 fig. 7), and on Salisbury Plain, where elongated fields up to 200 × 50 m seem to have been made by removing earlier cross-divisions (McOmish et al. Reference McOmish, Field and Brown2002, 56). While patterns of elongated fields between reave-like boundaries might reflect patterns of land allotment, for example, the need to equitably share out different resources available on a high(er) ground–low(er) ground axis, field morphology could also reflect their use (see below).

Chronology

Another aspect common to most of these field systems is the critical problem of dating them. As the ditches were mostly ‘outfield’ boundaries, seemingly located away from contemporary settlement or other focal points of activity, few artefacts found their way into them. In any case, the generally acidic sand soils hinder preservation of organic materials that could be radiocarbon dated. Moreover, where cultural material is present in ditch fills, it is more likely to reflect them ceasing to be maintained and going out of use than indicating when they were constructed. Stratigraphic relationships between ditches were frequently indistinct owing to the homogeneity of fills, which also impacts on the visibility of recuts and other evidence for ditch maintenance.

At most of the sites discussed here, initial excavation of spaced slots was followed – with selected ditches – by manual turning over and sifting of the fills along the intervening stretches, in some cases for lengths of 10–20 m or more (Fig. 8.1), specifically to search for artefacts. Despite this, quantities of finds remained low, and dating relies heavily on those instances where, by design or chance, excavation identified focal points in the field systems: structures, occupation areas, burials, or placed deposits within ditches. In this light, the sort of supervised machine excavation of prehistoric field ditches advocated by the East of England Research Framework has much to recommend it as a way of achieving large samples and maximising recovery of artefactual/environmental evidence (ALGAO 2019).

Paradoxically though, the absence of artefacts in ditches to some extent helps reinforce a prehistoric date: field ditches from later, materially richer, ages (for example, the Roman or medieval periods) usually contain at least some abraded pottery, ceramic building material, or other settlement debris, originally deposited via the common practice of indirect manuring of fields with midden waste (Yates Reference Yates2007, 81). Of course, later prehistoric farmers did this too (Bradley Reference Bradley1978, 41; Bell Reference Bell1983; Parker Pearson Reference Parker Pearson, Champion and Collis1996, 125–7; Rudling Reference Rudling2002, 203–38, 249–51; Guttman et al. Reference Guttmann, Simpson, Davidson, Brickley and Smith2005, 68, 73–4; Bradley & Yates Reference Bradley and Yates2007, 97; Yates Reference Yates2007, 138; Nowakowski Reference Nowakowski, Allen, Sharples and O’Connor2009, 120) but the friable, bonfire fired pottery of the period survives poorly, particularly if originally deposited on the ground surface. Indeed, this is a plausible origin for the uniformly small, abraded later Bronze Age–Early Iron Age-type potsherds found at a low density (and without any later material) throughout many of the excavated field ditches.

The alternative, based on its limited quantity and poor condition, is to view this cultural material as residually deposited in later ditches that happened to cut through Late Bronze Age–earlier Iron Age features or surface deposits. However, unlike some areas that were excavated for the East Anglia One windfarm (discussed below), at most of the sites described here (excepting Brantham) there was little sign of settlement that could be the source of the ditch material. Furthermore, truncation of individual settlement features would likely result in localised concentrations of residual finds in sections of ditches rather than the uniform, low density distribution throughout the ditch systems seen at sites like Trimley, Felixstowe Academy, Ravenswood Area ‘T’, and Martlesham.

If this interpretation of the origin of the ditch associated cultural material is correct then it would raise important possibilities regarding the longevity of the field systems: they may still have been actively manured into the Late Bronze Age and potentially beyond, though the low diagnostic value of the pottery precludes chronological precision. Of course, this says more about the later use of relict (Middle?) Bronze Age fields than their original function, and there is a possibility that, although the land was still being cultivated (or grazed), the remnant boundaries no longer meaningfully defined the limits of agricultural plots. In this sense, there could have been a reversion to something like the pre-field system landscape of the earlier Bronze Age, in which land was farmed – and tenure understood – without entrenched boundaries physically defining plots (Knight & Brudenell Reference Knight and Brudenell2020, 206–8, 226, 237): the short-lived 2nd millennium bc ‘experiment in (hu)man/land relations’ (Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009b, 256) may have been over.

This naturally leads on to questions of overall chronology. While the evidence from the west of the eastern region, particularly Cambridgeshire, overwhelmingly places large scale field system construction in the Middle Bronze Age (Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009a, 55; Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009b, 252–3), there are Late-attributed field systems in the south, notably South Hornchurch, which apparently saw three phases of development during the c. 10th–8th centuries bc, associated with both plainware and decorated PDR pottery (Guttman & Last Reference Guttman and Last2000). The early 1st millennium bc (phase I.2–3) ditched land divisions at North Shoebury may be another Thames Estuary example (Wymer & Brown Reference Wymer and Brown1995, 11, 20–33). Depending on the overall organisation and scale of these Suffolk sub-divided landscapes (see below), the evidence from Wessex, the North and South Downs and, perhaps, some Fenland sites such as Welland Bank, that co-axial reave-type strips (Middle Bronze Age) pre-date aggregate patchworks (Late Bronze Age?) (Field Reference Field2001, 59; McOmish et al. Reference McOmish, Field and Brown2002, 56; Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009a, 65; English Reference English2013, 3, 134) may be relevant, though this sequence was ostensibly reversed at Heathrow Terminal 5 (Leivers Reference Leivers, Lewis, Leivers, Brown, Smith, Cramp, Mepham and Phillpotts2010, 136), and at Clay Farm the sequence of strip-fields superseded by rectilinear fields/enclosures, and then by settlement within those enclosures, all took place within the Middle Bronze Age (Mortimer & Phillips Reference Phillips and Mortimer2012, 19–23).

As highlighted at the outset, the dating of almost all the south-east Suffolk field systems described here is very coarse and in the better cases (eg, Martlesham) generally only provides a later Bronze Age terminus ante quem. It certainly seems likely that the abraded PDR pottery, where present, relates to later rather than primary use of the fields. A picture of Late Bronze Age activity taking place within field systems that had – in all probability – been established in the Middle Bronze Age, but without much, if any, evidence for maintenance/alteration of the early field ditches, would be in keeping with a pattern seen widely elsewhere. The preliminary, but clearly better, dating evidence from Walton Green, combined with the complete Middle Bronze Age urn deposited in a related boundary at Felixstowe Academy, appears to present a more ‘typical’ sequence of a Middle Bronze Age sub-divided landscape that may have remained open/in use, to some extent, into the later Bronze Age.

What, though, should be made of the evidence from Game Farm, Brandon (Fig. S20)? Confronted by the discrepancy between the Middle Bronze Age radiocarbon dates and large c. 9th–8th century bc pottery assemblage from the structures, something has to give. Given the potential age/stratigraphic insecurity of oak charcoal from surface deposits, the radiocarbon dates seem questionable. By contrast, the stratigraphy of the field ditches cutting through the surface deposits of several of the Late Bronze Age structures appears unequivocal – the ditches must have been dug, or at least recut, in the Late Bronze Age or after. There may be a hint here that some of Suffolk’s Bronze Age field systems have similarities with those of Essex as much as Cambridgeshire.

Orientation and scale

The most striking characteristic of the field systems described here is their consistently broadly rectilinear layout, with the principal axes always aligned north-west to south-east by north-east to south-west or slight variations thereof. On sites with any degree of slope, principally Easton and to some extent Martlesham, these orientations respect the contours. However, where relief is not pronounced and the immediate topography does not predispose any particular alignments – as with the field systems to the south of Ipswich and on the Felixstowe peninsula – the widespread establishment of land divisions on north-west to south-east by north-east to south-west orientations may have referenced the main features in the wider physical landscape: the north-west to south-east landform of the peninsula itself, and the positions of the rivers Orwell and Deben. Unsurprisingly, where the field systems are directly adjacent to rivers/streams, their axes respect the positions of the watercourses, as at Game Farm, beside the Little Ouse. The impression is of fields laid out sympathetically with their landscapes rather than in the ‘terrain oblivious’ fashion of some Bronze Age land divisions (Yates Reference Yates2007, 134–5; Fleming Reference Fleming2008, 196; Green & Gosden Reference Green, Gosden, Gosden, Green, Cooper, Creswell, Donnelly, Franconi, Glyde, Kamash, Mallet, Morley, Stansbie and ten Harkel2021, 235).

The broad trackway at Trimley (Fig. 2) may head down towards the King’s Fleet and marshland (summer grazing?) along the south side of the Deben, possibly reflecting practical concerns and the need to link field systems to other landscape resources required by people or livestock. The possible trackway bounding the west side of the field block at Brantham (Fig. 2) might similarly lead downslope towards marshes/grazing along the north bank of the Stour, while that at Leiston (Fig. 7) is aligned down a shallow valley leading to the Hundred River. The orientation of the Leiston trackway dissects the east–west coastal promontory, between the Hundred River to the south and marshland to the north, that Leiston occupies, with the identified field system and associated farmstead enclosures on the higher ground approximately midway between the two; this is similar to the (low) ridge edge/top positions of Brantham, Trimley, and Felixstowe Academy, from where the main axial trackway of Walton Green Field System 2 also headed downhill towards marshes beside the river Orwell.

But at both Trimley and Martlesham (and possibly several of the other sites described here) spatial relationships between trackway/boundary alignments and ring-ditches/barrows in their immediate landscapes highlight how ostensibly less practical considerations, such as the positions of earlier monuments, might also influence the layout of land divisions and routeways. Conceivably, earlier burial grounds and favoured habitation locales (such as Martlesham Area 2; Fig. 5; Fig. S14) might have formed nodal points around which the transition from an open landscape and possibly partly mobile way of life to a bounded and increasingly sedentary agricultural existence was framed. Relationships between later Bronze Age boundaries and earlier monuments have been widely noted elsewhere, and there are many ways of interpreting the phenomenon (Cooper Reference Cooper2016). Deliberate positioning of boundaries in relation to earlier barrows or settlement areas might, for example, have been intended to solicit the assistance of forebears in guarding and controlling the land and its productivity (see Wiseman et al. Reference Wiseman, Allen and Gibson2021, 729–31), to harness the ‘symbolic capital’ (Cooper & Edmonds Reference Cooper and Edmonds2007, 76, 133) of monuments and the ancestors, real or imagined, with whom they were associated in order to legitimise claims to land (Johnston Reference Johnston2005, 18; Yates Reference Yates2007, 134; Knight & Brudenell Reference Knight and Brudenell2020, 208), or to connect in some other way with the perceived ancestral/supernatural forces therein – Fleming’s ‘monumental directionality’ (Reference Fleming2008, 198).

However, if the earth-fast land divisions of the later Bronze Age simply gave physical form to patterns of land use and tenure that already existed, and which were expressed on the ground in the positioning of houses, barrows, and other monuments, then the coincidence of ditched boundaries with these earlier foci just reflects those pre-existing patterns – the ‘operational grain’ of the landscape – being ‘entrenched’ (Knight & Brudenell Reference Knight and Brudenell2020, 206–10; cf. Johnston Reference Johnston2005, 11–13, 17). This might explain the stratigraphically early (and not necessarily burial related) ring-gullies underlying the trackway ditches at Ravenswood Area ‘T’ and the similar ring-gully beside Ditch 46 at Martlesham: they could have been markers of tenure/land allotment within the pre-field system landscape. One of the main axial alignments of the field system at Martlesham is neatly framed between the barrow group to the south-west and small ring to the north-east. A small ring-gully, without accompanying interment, at Colne Fen, Earith, Cambridgeshire, was similarly suggested as a possible landscape marker (Evans et al. Reference Evans, Brudenell, Patten and Regan2013, 95 fig. 4.2, 100–1, 128).

Relationships of some Suffolk field systems with earlier monuments might therefore reflect similar processes of entrenchment. But at some other sites, notably Flixton (Fig. S19), the influence of earlier monuments is less clear. While one of the later prehistoric field ditches there (Ditch 9) cut directly across an earlier Bronze Age barrow that was still upstanding, and another (Ditch 15) cut Early Bronze Age Ring-ditch 069 but kinked markedly to avoid its still-extant mound, many of the other, numerous, barrows and monuments in this landscape seem not to have had any particular impact on field layout (Boulter Reference Boulter2022, 64–6, fig. 3.45). Although the date of the Flixton field system is uncertain, the arrangements there bear comparison with Barleycroft Farm/Over, where, on the west side of the river Great Ouse, earlier ring-ditches occupied nodal points in the Middle Bronze Age field system and the ditch boundaries ‘spun’ on them, but on the east bank of the river the relationship appears less sympathetic and more a matter of ‘accommodation’, with the field axes ‘boxing’ the barrows in (as described by Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009a, 53–5, figs 2.19–20). Similarly, Bronze Age field systems on the Wessex chalk downland often show only limited respect for existing barrows (Bradley Reference Bradley1984, 107; Reference Bradley1998, 149). Therefore, while some earlier monuments were respected/referenced, in other instances deliberately slighting, enclosing (and thereby controlling?), or overtly ignoring monuments that had marked and expressed earlier people/land relations may have been just as much parts of the process by which existing patterns of land tenure were (variously) re-organised or entrenched. Laying out field systems involved a process of compromise and accommodation with the past (Bradley Reference Bradley2002, 78–9), which gave rise to ‘A specific set of grammars…in terms of how Middle Bronze Age land boundaries intersect …with round barrows’ (Cooper Reference Cooper2016, 674, 679 fig. 7).

Leaving aside the impact of the particular natural and human geography of this landscape, it can also be noted that the same (north-west to south-east by north-east to south-west) axial alignments are widely found across Bronze Age field systems elsewhere in the country, often at odds with topography (Fowler Reference Fowler2000, 25; McOmish et al. Reference McOmish, Field and Brown2002, 54–5, fig. 3.4; Yates Reference Yates2007, 136; English Reference English2013, 134; Evans et al. Reference Evans, Brudenell, Patten and Regan2013, 148; Bradley Reference Bradley2019, 219). This preferential alignment might have held cosmological significance, perhaps relating to sunrise/sunset at the solstices or the position of the sunrise more generally (Yates Reference Yates2007, 136; Green & Gosden Reference Green, Gosden, Gosden, Green, Cooper, Creswell, Donnelly, Franconi, Glyde, Kamash, Mallet, Morley, Stansbie and ten Harkel2021, 248–51). A preference for south/south-east facing orientations, possibly reflecting similar concerns, is apparent in other aspects of later prehistoric life, for example, the architecture of barrows and positioning of round-house entrances (Bradley 1998, 149–50; Reference Bradley2002, 77; English Reference English2013, 144; Green & Gosden Reference Green, Gosden, Gosden, Green, Cooper, Creswell, Donnelly, Franconi, Glyde, Kamash, Mallet, Morley, Stansbie and ten Harkel2021, 255).

Their regular, broadly ‘grid’-like, appearance raises the important question of whether these field systems were parts of a large scale, perhaps planned, co-axial landscape extending over tracts of the south-east Suffolk coast. Alternatively, they might have been self-contained (‘aggregate’: Bradley Reference Bradley1978, 268–9; Yates Reference Yates2007, 15; Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009a, 46, 64–5) patchworks of fields and enclosures laid out in more ad hoc fashion by individual farmsteads/small communities, albeit linked by trackways to neighbouring settlements and resources available in other parts of the landscape (for example, areas of communal grazing or woodland). Cropmarks in the area have previously been interpreted as tending towards the latter interpretation (Hegarty & Newsome Reference Hegarty and Newsome2005, 37) but the regular appearance of some of the extensive cropmarks in and around Martlesham (Fig. 9) could perhaps indicate the presence of wider areas of planned landscape, as may be true of parts of the Norfolk coast (Albone et al. Reference Albone, Massey and Tremlett2007, 43).

Fig. 9. Cropmarks around Martlesham (cropmark data kindly supplied by Suffolk Historic Environment Record and Historic England National Mapping Programme)

Fig. 10. Sites discussed in the text

Extensive excavation of Bronze Age sub-divided landscapes elsewhere, for example, the west Cambridgeshire Fen edge, have found major differences between them, notably in the presence/absence of droveways but also in field shape, size, and alignment (Yates Reference Yates2007, 84–97; Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009b, 241–2; Knight & Brudenell Reference Knight and Brudenell2020, 146, 390). Mark Knight and Matthew Brudenell (Reference Knight and Brudenell2020, 390) note that ‘consistency only occurred in blocks … open-up a big enough space and the configuration of a particular fieldsystem is liable to change … On current form, there would look to be as many types of fieldsystems as there are sites’. This underlines the huge variation across and between Bronze Age field systems, the frequent impossibility (or usefulness) of straightforward categorisation, and the need to investigate large areas: the exposures of Suffolk field systems to date are too small to properly understand their scale and structure.

The nature of agriculture and landscape context

As noted, despite extensive bulk sampling, the presence/survival of plant macrofossils and animal bone was almost non-existent across these field systems. In consequence it is impossible to know how they were used except indirectly, by inference from their topographical locations, soils and layouts – an approach not without pitfalls (Cooper Reference Cooper2018, 8). Slightly more information comes from those focal points where structures were identified, although there are frequently questions over their contemporaneity with the fields they were set in.

All but two of the sites share light sand and gravel geology: either Lowestoft Formation or Kesgrave Subgroup glacial deposits, overlying coarse, shelly sands and gravels of the Red Crag Formation (Table 2; Williamson Reference Williamson2005, 3–4; British Geological Survey 2024). Easton’s setting is more complex, being on a slope with head deposits at its foot and glacial boulder clay, the edge of the High Suffolk clay plateau, to the north and east. Leiston sits on fairly light glacial clay/silt overlying the Crag sand. Soils are typically free-draining, acid and sandy (Ipswich environs and Martlesham), loamy (Felixstowe, Trimley, and Brantham), or base-rich (Leiston) in composition (Cranfield University 2024). From at least the Anglo-Saxon period until the early 20th century the sandy area east of Ipswich and Woodbridge was predominantly lowland heath, supporting acid grassland, ling, gorse, and broom, and suitable primarily for grazing sheep (and later rabbits) due to its low fertility, excessive drainage, and shallow rooting zone (Armstrong Reference Armstrong1971). This propensity towards sandy soils, also apparent in the Breckland location of Game Farm, Brandon, might reflect the limited technology available to Bronze Age farmers; that is, their reliance on breaking ploughs or ards, without coulters or mouldboards to slice and turn over heavy soils and thereby help them dry out (Payne Reference Payne1957, 74–9; Williamson Reference Williamson2005, 9). In a period of relatively low population density, communities may also have lacked the workforce with which to attempt to alleviate the poor drainage of East Anglia’s clays.

Table 2. Topographical contexts of the bronze age field systems

However, the extent to which this was marginal land can be overstated. Part of the reason for the impoverished state of the Sandlings and Breckland soils is past human activity: tree clearance followed by over-grazing or cultivation by prehistoric farmers is thought to have led to leaching of nutrients and formation of podzols (Dimbleby Reference Dimbleby1962; Armstrong Reference Armstrong1971, 17–18; Rackham Reference Rackham1986, 286–91; Williamson Reference Williamson2005, 50; see also English Reference English2013, 10). The agricultural value of these soils in the Bronze Age, presumably soon after land was first cleared, might have been rather better than it was by the medieval period, perhaps particularly under a low intensity, shifting cultivation regime (eg, a short-fallow system; Barrett Reference Barrett1994a, 143) or if they were used primarily for grazing animals whose dung could help to replenish soil nutrients (Williamson Reference Williamson2005, 56–7; Serjeantson Reference Serjeantson2007, 83).

Furthermore, the Felixstowe and Shotley peninsulas are blanketed with aeolian loess deposited under periglacial conditions during the Late Devensian and subject to varying degrees of subsequent weathering and reworking (Catt Reference Catt and Shotton1977; Reference Catt, Limbrey and Evans1978; Eden Reference Eden1980; Ballantyne & Harris Reference Ballantyne and Harris1994, 160; Scheib & Lee Reference Scheib and Lee2010, 1). The fine silt subsoil, 0.4–0.5 m thick, encountered at Felixstowe Academy and Trimley derives from these loess deposits. In places the Bronze Age ditches could be seen cutting the subsoil; in others their silt fills were indistinguishable from it. Loess soils are among the most fertile in the world, providing good water retention and aeration, extensive penetration by roots, easy cultivation, and a high mineral and nitrogen content, though they are also at high risk of structural breakdown from overworking and erosion (Catt Reference Catt, Limbrey and Evans1978, 17; Reference Catt2001). Chemical analyses suggest that prior to weathering and human activity the distribution of loess soils across East Anglia was more extensive (Scheib & Lee Reference Scheib and Lee2010, 5–6). It may be that some of the other sites discussed here still had, in the Bronze Age and perhaps earlier Iron Age, at least a thin covering of loess, and that the Bronze Age field systems were preferentially sited on these high grade soils (see Yates Reference Yates2007, 137–8).

At many of the sites there was a reduction in archaeologically visible activity over time. Land that had been sub-divided into fields, sometimes with suggestions of long use, during the latter part of the Bronze Age appears to have been abandoned or at least less intensively used by the Early/Middle Iron Age. This could be a consequence of soil degradation, in which case the evidence for Late Bronze Age manuring with midden material might represent an attempt to fertilise and bulk-up depleted soils. However, wider social and economic factors may well have been involved as apparent cessation of field system construction/maintenance by the end of the Bronze Age is a near-universal phenomenon.

This brief review shows that the field systems were sited on light land that was in some cases very fertile and in others not necessarily as impoverished as it was by later periods. Turning to the evidence from the sites themselves, there is a clear difference in morphology between two components of the sub-divided landscape at Martlesham. The set of squarish enclosures in the north-west are too small to have been used for any duration as pasture for livestock; nor does their size fit well with arable cultivation. Moreover, aspects of their layout, particularly the trackways funnelling into some of them, presumed hedged boundaries surrounding them, and, where evidence survived, their narrow corner entrances (see Pryor Reference Pryor2001, 417), all suggest a connection with temporarily corralling livestock, perhaps for sorting, inspection, milking, shearing, branding, or slaughter. This bringing in of animals logically suggests an ‘infield’ location close to settlement. The larger ditched fields in the south-east may have been the outfield where some of these livestock normally grazed. These might have been sheep rather than cattle given the dry soil and sort of heathland vegetation it typically supports (Williamson Reference Williamson2005, 56), though the above points regarding the Sandlings’ ‘original’ soil quality should be borne in mind.

Small amounts of barley were found in the ring-gully at Martlesham and two contexts at Felixstowe Academy, but these grains were neither well-dated nor securely stratified and, in the absence of larger cereal/chaff assemblages, could have come from elsewhere. Barley found in Middle Bronze Age features at Red House Lane, Leiston, is more contextually secure, as are barley and both hulled and unhulled wheat from Walton Green, where chaff from cereal processing was also present (Adams Reference Adams2018; Clarke Reference Clarke2023, 409). Possible loomweights in two of the structures at Ipswich Academy imply some sheep farming. The sub-divided landscape at Game Farm was interpreted as primarily associated with stock, based (in the absence of surviving faunal remains) on the layout of the enclosures and possible droves; loomweights and spindle-whorls also suggest the presence of sheep there, while a very few emmer/spelt grains from the round-houses could show that those crops were grown (Gibson Reference Gibson2004, 53, 56–7; Scaife Reference Scaife2004, 47). Clay loomweights and a possible spindle-whorl were also recovered from the Late Bronze Age settlement beside the field system at Flixton, where small quantities of cereal grains and the presence of numerous four-/six-post raised(?) structures also attest to stored crops (Boulter Reference Boulter2022, 149).

The considerable width (c. 16 m) of the corner entrance to the field at Easton also makes best sense in relation to livestock; cattle could have grazed in wet meadow beside the Deben, 250 m downslope. There was no direct evidence for the nature of Bronze Age agriculture at Easton but the Early Iron Age inhabitants, who may still have used the field(s), certainly kept sheep, cattle, and pigs, with a possible loomweight suggesting the former were exploited partly for wool; burnt but unidentifiable cereal grains and a single pea (Fabaceae sp.) suggest mixed farming. Similar mixed farming, including livestock grazing in the relict fields, and cultivation of emmer, naked wheat, and oats, took place at Early Iron Age Brantham, although this again says nothing about the primary use of the field system. Trackways at Brantham (where the fields also had wide corner entrances suitable for cattle), Trimley, Walton Green, Ravenswood Area ‘T’, Martlesham, and Leiston might be droves for moving livestock, but access to and from arable fields, and longer distance transport through sub-divided landscapes for communities living either side, would also be important (Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009b, 246). The considerable width of the track at Trimley (10–13 m) and, to an even greater extent, central Trackway 2 at Martlesham, which widened as it approached the excavation area (18–22 m), are in the same order of magnitude as the ‘great’ droveways at Colne Fen, Earith, which were interpreted as components of a cattle handling and paddock system, though possibly primarily for use by through-traffic (Evans et al. Reference Evans, Brudenell, Patten and Regan2013, 96–7, 103, 111–12, 128–31, 148–50). Whether they were intended mainly for local or longer distance movement, construction of trackways implies a need to keep livestock/carts off land on either side – either it was under crops or was grazed by animals that had to be kept separate.

The few cereal grains at sites like Game Farm and Red House Lane and the larger assemblages from Walton Green and excavations of Bronze Age settlements in Suffolk (see below), show that cereal crops were grown, while the presence of probably manure derived potsherds suggests that some of the sub-divided land parcels described here were (at least in later times) cropped. However, lack of clear evidence for field sizes makes it difficult to reconstruct how they might have functioned in an arable context. Theoretically (and despite the presence of potential pastoral indicators, such as trackways and wide/corner entrances) the strip fields at Leiston and Brantham, with their relatively long, narrow shapes, could lend themselves to strip ploughing. Although their exposed portions (max. 99.8 m at Leiston) fall some distance short of the typical c. 200 m (220 yd) furlong that a team would plough before resting/turning in medieval open fields (Rackham Reference Rackham1986, 168), smaller prehistoric draft oxen and plough-teams might have reduced this distance. It is interesting, then, that the more complete Bronze Age strip-type fields seen elsewhere have lengths about 160–200 m. However, this similarity is probably coincidental: medieval selions were a product of the mechanics of working with a heavy, mouldboard plough and large team required to pull it, which needed considerable turning space and made it logical to plough for as long as possible in one direction before turning (Banham & Faith Reference Banham and Faith2014, 50–4, 70–2). Use of ards on these light soils would not necessarily give rise to this sort of ploughing pattern or predispose this field shape.

Indeed, early cultivation in Britain, as across north-west Europe, seems mainly to have involved cross-ploughing, a technique that was appropriate for light soils and relatively warm, dry climatic conditions, and which served to stir/aerate the soil while minimising evaporation (Payne Reference Payne1957, 77; Fowler & Evans Reference Fowler and Evans1967, 293–5; Rees Reference Rees1979, 38–9, 79–91; Fowler Reference Fowler1983, fig. 51; Fleming Reference Fleming2008, 190). Cross-ploughing can be carried out more easily in fields of approximately equal length and breadth and, while the probable small (sub-1acre/0.4 ha) square fields at Ipswich Academy (where a couple of isolated examples of possible plough/ard marks aligned with the Bronze Age field boundaries were recorded; Fig. S6) ostensibly appear less than ideal for efficient use and turning of an oxen-pulled ard, the larger land units partially exposed at Felixstowe Academy or perhaps Flixton could have been more practical. British evidence for Neolithic and Bronze Age cross-ploughing survives where conditions permit, for example, under sand dunes at Gwithian, Cornwall (Megaw Reference Megaw, Burgess and Miket1976, fig. 4.1; Nowakowski Reference Nowakowski, Allen, Sharples and O’Connor2009, 119–20), and beneath the long barrow at South Street, Avebury, Wiltshire (Ashbee et al. Reference Ashbee, Smith and Evans1979, 282–3, fig. 36). But it is not entirely clear whether these are representative of normal cultivation: plough marks beneath barrows might be ritualised preparation of the ground to receive the dead (Rowley-Conwy Reference Rowley-Conwy1987; Pryor Reference Pryor1998, 148) or connected with initial clearance and preparation of land for farming rather than the regular process of tillage (Ashbee et al. Reference Ashbee, Smith and Evans1979, 296). Furthermore, the scale of cereal cultivation in the British Bronze Age is still far from clear, and it may be that Francis Pryor and Andrew Fleming were right to see it primarily taking place at a horticultural level close to houses (Pryor Reference Pryor1998, 82, 118; Fleming Reference Fleming2008, 135). If so, preparing soil may normally have involved manual tillage with spades and hand tools (Rees Reference Rees1979, 6), in very small plots like those near the structures at Ipswich Academy, or those in the suggested field grid at Game Farm.

No doubt there was variation in the use of the different field systems described here but, taken as a whole, the limited combined evidence suggests that the sub-divided landscapes were used for mixed farming involving both livestock, perhaps predominantly sheep on the sandiest soils like those around Martlesham, and cultivation of barley and other cereal crops. Use for mixed farming would fit the evidence from excavated Bronze Age settlements in Suffolk. Middle Bronze Age midden deposits preserved in the top of a mine shaft at Grimes Graves, for example, showed that cereal crops (6-row barley and emmer) were being grown, possibly on soil made less acidic by upcast chalk from the mining, alongside a combination of cattle dairying and sheep farming, the latter mainly for meat (Legge Reference Legge and Mercer1981; Reference Legge1992). At Middle Bronze Age Fordham Road, Newmarket, charred barley and hulled wheat (spelt/emmer) grains were found, and quernstones provide evidence for cereal processing; the faunal assemblage predominantly consisted of butchery waste from adult cattle (Faine Reference Faine2017; Fosberry Reference Fosberry2017; Shaffrey Reference Shaffrey2017). The earlier Bronze Age (c. 1700–1500 bc) occupants of West Row Fen kept cattle as well as sheep and pigs; emmer was the main crop, followed by spelt, hulled barley, and flax (Martin & Murphy Reference Martin and Murphy1988, 355–6). The Early Bronze Age inhabitants of Sutton Hoo grew barley, oats and wheat, with a mix of arable and pastoral land use inferred from the layout of ditches and fences, though the contemporaneity of those is uncertain (Carver Reference Carver1998, 97–8; Hummler Reference Hummler and Carver2005, 446). At most of these sites there was also evidence for some hunting and gathering of wild resources. Of course, chronology is important here, as some of these sites pre-date the beginnings of formal field system construction, a development that may have been linked to other changes in agriculture.

David Yates interpreted the Bronze Age field systems of lowland England as primarily associated with livestock farming (Yates Reference Yates1999, 163–7; Reference Yates2001, 65–6, 73–4, 78; Reference Yates2007, 120). In East Anglia, Francis Pryor identified Fengate’s Bronze Age fields as hosting a sheep-centred agricultural economy (Pryor 1996; Reference Pryor1998, 89–108; Reference Pryor2001, 420) – and further argued that cereal cultivation was not an important part of agriculture in Britain until the c. Middle Iron Age (Pryor Reference Pryor1998, 82, 118, 148). More recently, these pastoral- and specifically sheep-centric views have been challenged by Christopher Evans on grounds of morphology (eg, many Bronze Age field systems, such as Bradley Fen, entirely lack droves) and the makeup of the associated faunal assemblages (which invariably have higher frequencies of cattle). Moreover, Evans reminds us that pastoral economies do not necessarily require formal field systems, which are perhaps more likely to reflect investment in the land itself, through enclosure, tree and stone clearance, and enrichment via manuring/middening – all processes just as likely to relate to growing crops (Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009a, 63; Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009b, 243–52). Evidence from the large scale excavations at Heathrow Terminal 5 also provides a corrective to the pastoral-centric view of Thames Valley Bronze Age field systems: environmental evidence confirms that all stages of cereal production and use occurred on site and that cereal production was at least as important there as livestock (Leivers Reference Leivers, Lewis, Leivers, Brown, Smith, Cramp, Mepham and Phillpotts2010, 142–3, 209).

Furthermore, it is far from clear whether specialised pastoral economies existed in this time and place on any great scale: reliable stock rearing required stores of cultivated fodder to keep some animals through winter and maintaining productive cropland relied, in turn, on manure from livestock; the two were heavily inter-dependent. Pryor (Reference Pryor1998, 149) plausibly argued that there could have been specialisation on individual farms and trade/exchange between them providing what they did not produce for themselves but this would still require a landscape in which a sizeable proportion of farmers were primarily cultivators. Hence, the region’s Bronze Age field systems are likely to have hosted mixed farming regimes, and their pastoral component may have been over-emphasised (Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009b, 248; Bradley Reference Bradley2019, 225).

But just like pastoral farming, arable cultivation can be carried out, even on a large scale, on land without physical boundaries (as in medieval open fields), providing that fallows/harvest aftermath are grazed communally. Thus, the sub-divided agricultural landscapes – and generally small fields – seen here are most likely to imply different things taking place in close proximity at the same time, perhaps on land parcels owned or controlled by different people. This all tends towards a rather small scale and individualistic system, somewhat contra the communal stockyards of Fengate (Pryor Reference Pryor1996; Reference Pryor1998, 89–108; Reference Pryor2001, 415–20; Reference Pryor, Lane and Coles2002, 27; critiqued by Evans Reference Evans, Evans, Beadsmoore, Brudenell and Lucas2009b, 243–52 and Knight & Brudenell Reference Knight and Brudenell2020, 208–10) and perhaps also some of David Yates’s ideas about field systems being bound up in a prestige economy of conspicuous production and consumption (Yates Reference Yates2007, 113, 120–32, 135–6; Bradley & Yates Reference Bradley and Yates2007, 97).

These south-east Suffolk field systems are similar in many respects to those mapped and investigated along the Thames Valley and west Cambridgeshire Fen edge but one ‘typical’ feature that is missing is the waterholes, probably because the free-draining geology made them impractical. This absence was compensated for by the location of all the field systems within a few hundred metres of springs/streams (Table 2). While a connection is usually assumed between waterholes and livestock (eg, Brown Reference Brown1988, 295; Yates Reference Yates2001, 65; Reference Yates2007, 120, 136), water is obviously essential for both animal and plant life, and the free-draining Sandlings soils would have posed some difficulties for crops. Barley, as found in very small quantities here, and generally the main Middle Bronze Age crop (Bradley Reference Bradley2019, 224), is resilient and versatile, able to grow in quite poor soils, and relatively drought tolerant, although it yields poorly on sand (Bradley Reference Bradley1978, 32). A consequence of the absence of waterholes and other ‘wet’ features is a lack of opportunities for palaeo-environmental investigation and the better understanding of field use and landscape context it could bring.

Distribution and landscape/land use model

These excavations provide only small windows on landscapes that, in all cases, extend beyond the investigated areas. In effect they constitute a landscape characterisation exercise carried out via a series of random ‘keyhole’ samples rather than a designed research strategy. As such there is no reason to suspect the sites are atypical and it is possible that similar features might be uncovered if any comparable sized area on the light soils of the south-east of the county was subject to intrusive investigation. This area has extensive cropmarks of fields, enclosures, and trackways identified and plotted through projects such as the Historic England National Mapping Programme (NMP), Rapid Coastal Zone Assessment Survey (RCZAS), and Aggregates Levy Sustainability Fund (ALSF) (Hegarty & Newsome Reference Hegarty and Newsome2005, 30–2, 37–41; Hegarty Reference Hegarty2006, 1, 9, 16–18, 20–9, 47–8; Good et al. Reference Good, Hegarty, Plouviez and Rolfe2007, 81, 83–4, 88–9, 105). Few of these have been sampled through excavation and securely dated. In the absence of ‘ground truthing’, dating remains reliant on apparent spatial relationships with landscape features of known age and on morphological comparison with other sites. In practice the latter can result in a tendency to assign Late Iron Age or Roman dates to field systems with rectilinear layouts, probably because the creation of planned-looking landscapes is often still widely assumed to have been impossible without large scale mobilisation of labour and resources under elite or state control, and that this could only have been achieved during those periods. A further factor underlying the apparent scarcity of Bronze Age sub-divided landscapes in northern East Anglia was the relatively low number of large scale archaeological excavations that had taken place – at least until recently – compared with neighbouring counties, particularly Cambridgeshire and Essex. Therefore, rather than a genuine absence, the previous lack of evidence for Bronze Age field systems in Suffolk (and Norfolk) resulted from not looking for them and not always recognising them when found.

In view of the results discussed here, at least parts of the extensive cropmark field systems on the Shotley peninsula, Trimley peninsula (Fig. 4), and further north along the coast, for example, in Martlesham, Waldringfield, Newbourne, and Hemley (Fig. 9), are likely to have Bronze Age origins. In addition to the recent excavations at Walton High Street (SHER FEX 299 and 451) mentioned above, areas of these cropmarks were investigated during construction of the cable route for the East Anglia One (EA1) offshore windfarm. Excavations exposed extensive ditch systems corresponding with parts of the cropmark complexes, as well as identifying Bronze Age burials, ring-ditches, and structures (SHER MRM 172b, Nicholson & Lotherington Reference Nicholson and Lotherington2019; SHER MRM 173c, Thomas Reference Thomas2019; SHER WLD 069, Lloyd-Smith et al. Reference Lloyd-Smith, Monahan, Peachey and Summers2019a; SHER HMY 043, Lloyd-Smith et al. Reference Lloyd-Smith, Monahan, Peachey and Summers2019b; SHER HMY 044, Nicholson Reference Nicholson2019). Ongoing analysis suggests that many of the cropmark and other ditches that were ‘tested’ within the EA1 cable easement, which were phased in the initial assessment reports (above) as Bronze Age, actually contain mainly residual finds deriving from later Bronze Age ‘open’ settlement in their vicinities; they can therefore only be broadly dated as ‘later prehistoric to Roman’ (Jonathan Tabor, pers. comm.). Nevertheless, there is reportedly good evidence of Middle Bronze Age land division in some of the investigated areas, hence publication by the Cambridge Archaeological Unit and Museum of London Archaeology is eagerly anticipated. The evidence for Bronze Age field systems complements the high concentration of barrows and cropmark ring-ditches across this area (Martin Reference Martin, Lawson, Martin and Priddy1981, fig. 25, 75–7; Williamson Reference Williamson2005, 10). Similar evidence has been recovered from the Norfolk coast, where excavations for the Bacton to Great Yarmouth gas pipeline found that at least some of the extensive co-axial field systems mapped by the NMP have Bronze Age origins (Albone et al. Reference Albone, Massey and Tremlett2007, 40–9; Reference Albone, Massey and Tremlett2008, 33–4; NHER 12828, Bates & Crowson Reference Bates and Crowson2004).

The widespread presence of Bronze Age field systems in south-east Suffolk having been established, what of the rest of the county? As discussed previously, some of the largest excavations carried out in Suffolk to date, in advance of gravel extraction at Flixton, in the Waveney Valley (Fig. 10), revealed a multi-period landscape including an extensive early rectilinear field system. Its ditches post-date the Early Bronze Age and pre-date the Roman period but are otherwise undated; it may well have Middle Bronze Age origins – certainly it exhibits strong morphological similarities to some of the other Suffolk field systems discussed here (Stuart Boulter, pers. comm.; Boulter Reference Boulter2022, 63–6, 149–51). Hints of ditched enclosure(s), albeit undated, were present further east along the Waveney catchment, adjacent to the Late Bronze Age farmstead at Bloodmoor Hill, Carlton Colville (SHER CAC 042, Heard Reference Heard2013, 24 fig. 9, 47, 77).

To judge from Game Farm, Brandon, beside the Little Ouse, at least some of the river valleys in the west of the county also had Bronze Age field systems. Excavation certainly demonstrates their presence along the River Snail. At Fordham Road, Newmarket, early boundary features, both ditches and a line of Early Bronze Age ‘tree-pits’, were associated with a substantial and long lived enclosed settlement that saw a major phase of occupation during the Middle Bronze Age (SHER NKT 047, Rees Reference Rees2017). A few hundred metres south, at Windmill Hill, Exning, a large Middle Bronze Age boundary ditch has been excavated on the east facing slope extending down to the river (SHER EXG 099, Woolhouse Reference Woolhouse2012; SHER EXG 105, Cass Reference Cass2014); there are indications of another substantial Bronze Age enclosed settlement on the hilltop (SHER EXG 082, Caruth Reference Caruth2006; Craven & Brudenell Reference Craven and Brudenell2011). Later Bronze Age settlements and associated field systems have been recorded further downstream, in and around Fordham and neighbouring Soham, both into Cambridgeshire (Yates Reference Yates2007, 98–9; Connor & Mortimer Reference Connor and Mortimerforthcoming). The Kennett Valley, east of Newmarket, has a similar chalk landscape with signs of Bronze and earlier Iron Age occupation, including a large cropmark hilltop enclosure overlooking the ford over the river (SHER MUN 039, MUN 040, MUN 047) – this may be another prime location to look for evidence of Bronze Age fields.

This leads on to the Suffolk Fen edge, with its concentrations of Bronze Age metalwork (Martin Reference Martin, Dymond and Martin1999a, 38–9; Malim Reference Malim2001, 17–19), the logical context for which might arguably be an area of high population density and associated closely managed agricultural resources (see Pryor Reference Pryor1998, 140). Similar coincidences of high volumes of bronze deposition with intensive settlement activity and field construction are found on the Southend peninsula, Essex, and elsewhere along the Thames gravel terraces (Yates Reference Yates2001, 78; Reference Yates2007, 114). However, Colin Pendleton argued (Reference Pendleton1999, 89) that the apparent frequency of bronze finds on the Suffolk Fen edge relates primarily to recovery/survival biases arising from modern land use rather than particularly high population or intensive activity during the Bronze Age. The temporal disjunct between the Middle Bronze Age field system and Late Bronze Age metalwork deposition at Bradley Fen (Knight & Brudenell Reference Knight and Brudenell2020, 221, 390, 395) also calls into question whether concentrations of bronzes necessarily equate with areas of intensive settlement and agriculture. Nevertheless, the contrast between the west Cambridgeshire Fen edge, with its extensive Bronze Age field systems, and the eastern Fen edge, with its early settlement spreads, burnt mounds and metalwork, but lack of sub-divided landscapes, is striking.

Given intrusive fieldwork on a similar scale, would the Suffolk/Norfolk Fen edge and river embayments be revealed to have Bronze Age field systems comparable to those found on the west side of the Fen basin? Interestingly, the large excavations on the fen edge at RAF Lakenheath did find traces of stratigraphically early field ditches beneath the far more extensive Iron Age and Roman boundary systems, though these were difficult to track confidently and remain undated (Jo Caruth, pers. comm.). Furthermore, just a few kilometres north of Lakenheath and a little way upriver from where the Little Ouse emerges into the fen, the Middle–Late Bronze Age ditched field/enclosure system at Game Farm is more likely to be a fortuitous, localised survival, protected there by colluvium/windblown sand deposits, than an anomaly. It is precisely preservation rather than genuine difference in landscape organisation which explains the extreme scarcity of Bronze Age sub-divided landscapes on the Suffolk (and Norfolk) Fen edge. The prehistoric archaeology of the eastern Fen edge is generally much closer to the surface and exposed to damage from modern land use: drainage, intensive farming, and peat wastage (Silvester Reference Silvester1991, 136; Healy Reference Healy1996, 177–8). These processes ‘have progressively reduced the south-eastern fen-edge to the extent that very little survives of post-Early Bronze Age occupation’; ‘Prior to erosion, the dry eastern margins once contained traces of other later second millennium bc features including Middle Bronze Age fieldsystems and later Bronze Age settlement.’ (Knight & Brudenell Reference Knight and Brudenell2020, 393–5, fig. 7.9).

In the hinterland of the fen edge, the landscape rises to a (relatively) high, dry plateau away from the bisecting river valleys. Excavation of a 15 km tract through these ‘high Brecks’ during dualling of the A11 between Barton Mills and Thetford found no clear evidence of Bronze Age land divisions to complement the density of barrows in the area (Martin Reference Martin, Lawson, Martin and Priddy1981, fig. 25, 75–7; Reference Martin, Dymond and Martin1999a, 38–9). While more investigation is needed before drawing firm conclusions, a likely reason for the apparent absence of Bronze Age fields – in an area of light soils similar to the Sandlings – is the scarcity of water. The practical and symbolic significance of water in this dry landscape may be reflected by a cluster of Late Bronze Age un-urned cremation burials surrounding a small, seasonally wet, waterhole at Chalk Hall Farm, Elveden, sited in one of a few locales where a clay capping over the free-draining chalk enables groundwater to collect (SHER ELV 085, Lees et al. Reference Lees, Hinman and Stump2013, Woolhouse in prep.). Bronze Age land use on the Breckland plateau would appear from this largely negative evidence to have been non-intensive and geared around peripatetic grazing of livestock, utilising scattered features like this one to water animals, and with burials perhaps re-inforcing communities’ claims to use them. Settlements – and perhaps more intensively used, bounded, and managed agricultural land – may have been focused along the low ground of the fen edge, several kilometres to the west, and the river valleys a similar distance to the north and south, with livestock grazing in the fen during the drier, summer months. Just such a pattern of seasonal transhumance was suggested for the Early Bronze Age inhabitants of West Row Fen (Martin & Murphy Reference Martin and Murphy1988, 357).

East of the high Brecks, but still some distance from the Black Bourn Valley, further to the east, recent geophysics and trial trenching has identified a series of large (c. 0.5–0.8 ha), predominantly curvilinear compounds/enclosures surrounded by ditches up to several metres wide and over a metre deep, constructed during the Middle Bronze Age (SHER EUN 075, EUN 076 and FKM 085, Billington & Cox Reference Billington and Cox2021). Part of a probable Late Bronze Age sword blade was found in the upper fill of one of the ditches, while pottery, worked/burnt flint, and charred grain suggest some level of domestic and other activities. There are indications that the surrounding land has a broadly contemporary rectilinear field system demarcated by less substantial ditches. The complex’s interfluvial location might be explained by its proximity to the groundwater fed meres at Rymer Point, which historically provided a valuable source of water for livestock grazing the heath and which may have allowed Bronze Age activity to extend some distance away from the river valley (Dymond Reference Dymond and Munby1968, 23–4; Martin Reference Martin, Dymond and Martin1999a, 38).

Similar primarily pastoral Bronze Age land use may be suggested by the evidence from some other excavated sites further ‘inland’, including Marham Park, Fornham All Saints (SHER FAS 056, Green Reference Green2018 Footnote ³ ), located on a slope and ridge overlooking the river Lark, and Green Lane, Haughley (SHER HGH 058, Woolhouse & Pullen 2020; forthcoming), on the edge of the Gipping Valley/mid-Suffolk clay plateau. At both sites wells/waterholes and associated burnt flint mounds appear most likely to have been connected with livestock farming, perhaps including butchery and processing of animal carcasses and skins. A Bronze Age double-ditch boundary or droveway at Haughley and a similar ditched and banked ‘ranch’-type boundary or sheep walk at Marham Park could have helped to move and corral animals through what appear – based on pollen – to have been largely open landscapes on the valley sides, though there were some fragments of rectilinear ditched fields/enclosures at Marham Park that might be remnants of a wider-scale field system (Green Reference Green2018, 85 fig. 22; Abby Antrobus pers. comm.). There are chronological differences between these landscapes: the burnt mounds and wells at Marham Park primarily date to the Beaker period/Early Bronze Age, the field remnants are Middle Bronze Age, while the long ditch and bank boundary appears to be Late Bronze Age/Early Iron Age. Material from both the waterhole and droveway at Haughley returned Middle Bronze Age radiocarbon dates, though the lower fills of the waterhole were undated and it could have been constructed earlier.

There may have been more concerted enclosure/sub-division of land further down the valley sides, as suggested by excavations 600 m south of Green Lane, at Fishponds Way, Haughley. Here, on a sandy slope adjacent to a minor tributary of the river Gipping, were traces of ephemeral and sterile-looking rectilinear ditches, which, though essentially undated, certainly had the appearance of a fragmentary Bronze Age field system like those seen in the south-east of the county (SHER HGH 060, Mlynarska & Woolhouse Reference Mlynarska and Woolhouse2020, 34–9, 99–100, figs 7–8). There was abundant evidence for Neolithic and Early Bronze Age activity at the site and the ditched trackway at Green Lane heads directly towards the stream close to this site in the valley bottom. Use of a droveway to control movement implies a need to keep livestock from straying over the apparently open grassland/pasture to either side, suggesting either different ownership or the intention to separate the animals being driven from those grazing the hillside. The extensive cropmark complex on the floor of the Lark Valley, just east of Marham Park, might perhaps hold similar evidence for Bronze Age land allotment along the river margins but has, to date, seen very limited archaeological investigation.

As at Haughley, there are other hints of Bronze Age land divisions extending onto the fringes of Suffolk’s clay soils. Notable in this regard is Hales Farm Barn, Withersfield, where excavation in advance of house building identified part of a sub-rectangular enclosure surrounded by a large ditch and either enclosing or adjacent to an earlier findspot of a Late Bronze Age hoard (SHER WTH 011, Bales & Topham-Smith Reference Bales and Topham-Smith2002); the evaluation trenches contained evidence of a possible associated field system, comprising shallow gullies with contemporary pottery (Sommers Reference Sommers1998; Colin Pendleton, pers. comm.). The site is in the upper reaches of the Stour Brook Valley, on the margin of the south Suffolk clay. What appears to be part of another small settlement, dating to the Late Bronze Age/Early Iron Age, has been found on the boulder clay above the valley of the river Brett, at Red Hill Road, Hadleigh, associated with a possible droveway and ditched field boundaries on the same alignments (SHER HAD 061, Meredith Reference Meredith2004).

Perhaps unsurprisingly, until recently there was no evidence for Bronze Age sub-divided agricultural land on the poorly draining High Suffolk clay plateau itself. This mirrored the sparse distribution of other visible indicators of Bronze Age occupation, such as round barrows and ring-ditches, in the area (Martin Reference Martin, Lawson, Martin and Priddy1981, fig. 25, 75–7; Reference Martin, Dymond and Martin1999a; Williamson Reference Williamson2005, 9), though it should be noted that Colin Pendleton believed this pattern to be largely illusory, caused by destruction of sites through a long history of arable cultivation (Reference Pendleton1999, 89–90). A few years ago, Early–Middle Bronze Age features, including a burnt mound, pond, waterhole, and light scatter of associated pits and post-holes, were identified on the clay interfluve at Yaxley, associated with clear evidence for woodland clearance by the early 2nd millennium bc (Brudenell & Kwiatkowska Reference Brudenell and Kwiatkowska2022). More recently, and since this article was drafted, another burnt mound complex has been investigated on the clay plateau at Laxfield, there associated with a small exposed area of rectilinear enclosures/fields bounded by ditches that contain Middle Bronze Age pottery (SHER LXD 135, Fern Reference Fern2022; Richard Mortimer pers. comm.). There is evidence for Bronze Age activity in similar topographical/geological contexts elsewhere in the region, for example, on the north-west Essex boulder clay plateau (Guttman Reference Guttman2000), but evidence for Bronze Age land division in these sorts of locales – here on flat and poorly draining clay several kilometres inland from the nearest significant river valleys (the Blyth and Alde) – is unusual in the region and a first for Suffolk. Parts, at least, of the heavy clayland were therefore being utilised by the Early Bronze Age, perhaps primarily as pasture/browse for livestock and exploitation of other ‘woodland’ resources, and there may have been some associated settlement (Pendleton Reference Pendleton1999, 89–90; Martin Reference Martin, Davies and Williamson1999b, 52). In some areas, Middle Bronze Age land division extended even onto these higher, interfluvial clays.

These hints of Bronze Age pastoral land use in the claylands find a parallel in medieval and early post-medieval farming in this landscape. Before the introduction of field drains from the 18th century onwards, ‘High’ Suffolk – particularly its flatter, northern parts – were a noted dairying region on account of their poor drainage and limited suitability for arable cultivation, but reliably lush grass cover (Martin Reference Martin, Christie and Stamper2012, 227). Evidence from some clayland sites in neighbouring counties, for example, the Middle Bronze Age ‘Ormesby-type’ enclosure at Swan’s Nest, Swaffham, located on the edge of the central Norfolk boulder clay, also suggests a main emphasis on livestock farming rather than arable cultivation (White Reference White2022, 105–8). However, it would be premature to ascribe Bronze Age use of the Suffolk claylands an exclusively pastoral basis given the clear evidence from ‘heavy land’ sites in Cambridgeshire, such as Striplands Farm, Longstanton, that Late Bronze Age settlements there were engaged in mixed farming (Evans & Patten Reference Evans, Patten, Brudenell and Taylor2011, 41).

This is obviously a rapid survey, heavily biased towards sites with which the author is familiar. No doubt there are numerous further examples and archaeologists working in other professional organisations operating in Suffolk will be aware of important sites currently undergoing post-excavation analysis. Caution must clearly be exercised in drawing too much from unpublished work that has not always progressed beyond preliminary analysis. Nevertheless, a provisional picture begins to emerge of quite widespread Bronze Age sub-divided agricultural land on the light soils of Suffolk’s coast and river estuaries, continuing inland up those rivers and their tributaries, and also occurring, where water was available, on the chalk and sands in the west of the county. It is also becoming increasingly clear – as in areas such as Kent (Yates Reference Yates2007, 110) – that landscape sub-division, probably with a pastoral focus, was pushing onto the central clay, in some places even the poorly draining interfluves.