The importance of chronology is reasserted as a means to achieving history and a sense of temporality. A range of current methods for estimating the dates and durations of archaeological processes and events are considered, including visual inspection of graphs and tables of calibrated dates and the summing of the probability distributions of calibrated dates. These approaches are found wanting. The Bayesian statistical framework is introduced, and a worked example presents simulated radiocarbon dates as a demonstration of the explicit, quantified, probabilistic estimates now possible on a routine basis. Using this example, the reliability of the chronologies presented for the five long barrows considered in this series of papers is explored. It is essential that the ‘informative’ prior beliefs in a chronological model are correct. If they are not, the dating suggested by the model will be incorrect. In contrast, the ‘uninformative’ prior beliefs have to be grossly incorrect before the outputs of the model are importantly wrong. It is also vital that the radiocarbon ages included in a model are accurate, and that their errors are correctly estimated. If they are not, the dating suggested by a model may also be importantly wrong. Strenuous effort and rigorous attention to archaeological and scientific detail are inescapable if reliable chronologies are to be built. The dates presented in the following papers are based on models. ‘All models are wrong, some models are useful’ (Box 1979, 202). We hope readers will find them useful, and will employ ‘worry selectivity’ to determine whether and how each model may be importantly wrong. The questions demand the timetable, and our prehistories deserve both.

Forty-four radiocarbon results are now available from the Ascott-under-Wychwood long barrow, and are presented within an interpretive Bayesian statistical framework. Three alternative archaeological interpretations of the sequence are given, each with a separate Bayesian model. In our preferred model, pre-barrow occupation including small timber structures and a midden was followed by a gap long enough to allow a turfline to form. Cists and the primary barrow were then initiated and the first human remains inserted into the cists; there was subsequently a secondary extension to the barrow. In the Bayesian model for this interpretation, occupation goes back to the fortieth century cal. bc, the midden being quite short-lived in the latter part of the fortieth or first part of the thirty-ninth century cal. bc. The gap was very probably not less than 50 years long, in the latter part of the thirty-ninth century cal. bc and the first half of the thirty-eighth century cal. bc. The barrow was begun between 3760–3695 cal. bc and extended in 3745–3670 cal. bc, probably within a generation. The first bodies were inserted in 3755–3690 cal. bc, contemporaneously with the primary barrow, and the last remains were probably deposited in the 3640s or 3630s cal. bc. The use of the monument probably did not exceed three to five generations. In an alternative interpretation of the sequence, greater continuity is seen between the underlying timber structures and midden on the one hand and the cists on the other, which could have preceded the initiation of the barrow itself. The Bayesian model for this interpretation suggests the gap between occupation and barrow was much shorter, probably of only 1–40 years' duration. It gives slightly different other estimates for the sequence but agrees with the main model in suggesting an overall short span of use for the whole monument. In a third interpretation, some of the human remains are interpreted as older than the cists and barrow. The Bayesian model for this again gives slightly different estimates but suggests that such putatively ancestral remains would not have been more than a decade or two older than the initiation of cists and barrow. Results are briefly discussed in relation to the overall sequence from occupation and midden to monument, the brevity of monument use, and issues of remembrance.

Forty-four radiocarbon results are now available from the Hazleton North long cairn, and are presented within an interpretive Bayesian statistical framework. Three alternative archaeological interpretations of the sequence are given, each with a separate Bayesian model. In our preferred model the cairn is considered to be a unitary construction, following on from the pre-cairn midden and other activity after a short interval during which the site was cultivated; bodies of the recently dead were subsequently interred in the chambered areas. Further human remains were deposited in the entrances to the chambers slightly later in the Neolithic, after the primary phase of use of the cairn for burials. This model suggests that the cairn was constructed in the first half of the thirty-seventh century cal. bc, and that its primary use for burial lasted for only two or three generations, ending probably in the 3620s cal. bc. A second model which varies only in postulating continuity between the pre-cairn activity and the cairn itself has poor overall agreement, suggesting that this interpretation is improbable. The third model explores the possibility that some of the human remains (those where the deposition of intact corpses cannot be strongly inferred from the archaeological record) may have been curated for a considerable time since death when deposited in the tomb. This interpretation suggests a slightly later date for the construction of the cairn, in the middle decades of the thirty-seventh century cal. bc, and suggests that any human remains which were not interred as corpses were less than a century old when deposited. The correspondence between the bones most likely, on chronological grounds, to be ‘ancestral’ and those most likely, on archaeological grounds, not to have been deposited as intact corpses is, however, poor. For this reason we feel there is no clear evidence that the human remains at Hazleton were not deposited shortly after the deaths of the individuals concerned, and we prefer model 1.

Twenty-seven radiocarbon results are now available from the Fussell's Lodge long barrow, and are presented within an interpretive Bayesian statistical framework. Three alternative archaeological interpretations of the sequence are given, each with a separate Bayesian model. It is hard to decide between these, though we prefer the third. In the first (following the excavator), the construction is a unitary one, and the human remains included are by definition already old. In the second, the primary mortuary structure is seen as having two phases, and is set within a timber enclosure; these are later closed by the construction of a long barrow. In that model of the sequence, deposition began in the thirty-eighth century cal. bc and the mortuary structure was extended probably in the 3660s–3650s cal. bc; the long barrow was probably built in the 3630s–3620s cal. bc; ancestral remains are not in question; and the use of the primary structure may have lasted for a century or so. In the third, preferred model, a variant of the second, we envisage the inclusion of some ancestral remains in the primary mortuary structure alongside fresh remains. This provides different estimates of the date of initial construction (probably in the last quarter of the thirty-eighth century cal. bc or the first half of the thirty-seventh century cal. bc) and the duration of primary use, but agrees in setting the date of the long barrow probably in the 3630s–3620s cal. bc. These results are discussed in relation to the development and meanings of long barrows at both national and local scales.

Thirty-one radiocarbon results are now available from the West Kennet long barrow, and are presented within an interpretive Bayesian statistical framework. Two alternative archaeological interpretations of the sequence are given, each with a separate Bayesian model. In our preferred interpretation, the barrow is seen as a unitary construction (given the lack of dating samples from the old ground surface, ditches or constructional features themselves), with a series of deposits of human remains made in the chambers following construction. Primary deposition in the chambers is followed by further secondary deposition of some human remains, including children, and layers of earth and chalk, the latest identifiable finds in which are Beaker sherds. In the Bayesian model for this sequence, the construction of the monument at West Kennet, as dated from the primary mortuary deposits, occurred in 3670–3635 cal. bc, probably in the middle decades of the thirty-seventh century cal. bc. The last interments of this initial use of the chambers probably occurred in 3640–3610 cal. bc. The difference between these two distributions suggests that this primary mortuary activity probably continued for only 10–30 years. After a hiatus probably lasting for rather more than a century, the infilling of the chambers began in 3620–3240 cal. bc and continued into the second half of the third millennium cal. bc. In an alternative interpretation, we do not assume that all the people dated from the primary mortuary deposits were placed in the monument in a fleshed or partially articulated condition; they could therefore have died before the monument was built, although they must have died before the end of the formation of the mortuary deposit. In the Bayesian model for this interpretation, the monument appears to belong either to the thirty-seventh century cal. bc or the mid-thirty-sixth century cal. bc, and deposition again appears short-lived, but the model is unstable. Results are discussed in relation to the setting and sequence of the local region.

Twenty-three radiocarbon results are now available from the Wayland's Smithy long barrow, and are presented within an interpretive Bayesian statistical framework. Four alternative archaeological interpretations of the sequence are considered, each with a separate Bayesian model, though only two are presented in detail. The differences are based on different readings of the sequence of Wayland's Smithy II. In our preferred interpretation of the sequence, the primary mortuary structure was some kind of lidded wooden box, accessible for deposition over a period of time, and then closed by the mound of Wayland's Smithy I; Wayland's Smithy II was a unitary construction, with transepted chambers, secondary kerb and secondary ditches all constructed together. In the Bayesian model for this interpretation, deposition began in the earlier thirty-sixth century cal. bc, and probably lasted for a generation. A gap of probably 40–100 years ensued, before the first small mound was constructed in 3520–3470 cal. bc. After another gap, probably of only 1–35 years, the second phase of the monument was probably constructed in the middle to later part of the thirty-fifth century cal. bc (3460–3400 cal. bc), and its use probably extended to the middle decades of the thirty-fourth century cal. bc. Results are discussed in relation to the local setting, the nature of mortuary rites and the creation of tradition.

Our final paper in this series reasserts the importance of sequence. Stressing that long barrows, long cairns and associated structures do not appear to have begun before the thirty-eighth century cal. bc in southern Britain, we give estimates for the relative order of construction and use of the five monuments analysed in this programme. The active histories of monuments appear often to be short, and the numbers in use at any one time may have been relatively low; we discuss time in terms of generations and individual lifespans. The dominant mortuary rite may have been the deposition of articulated remains (though there is much diversity); older or ancestral remains are rarely documented, though reference may have been made to ancestors in other ways, not least through architectural style and notions of the past. We relate these results not only to trajectories of monument development, but also to two models of development in the first centuries of the southern British Neolithic as a whole. In the first, monuments emerge as symptomatic of preeminent groups; in the second model, monuments are put in a more gradualist and episodic timescale and related to changing kinds of self-consciousness (involving senses of self, relations with animals and nature, perceptions of the body, awareness of mortality and attitudes to the past). Both more distant and more recent and familiar possible sources of inspiration for monumentalization are considered, and the diversity of situations in which mounds were constructed is stressed. More detailed Neolithic histories can now begin to be written.