• home
• search site
• academic papers
  • thesis
  • black death
  • buildings archaeology
  • dendro-chronology
  • GIS in archaeology
  • publications
  • statistics
  • devizes heads
• news
  • blog
  • upcoming events
  • related blogs
• photo archive
  • main archive
  • carpenters marks
  • hampshire crucks
  • scarf joints
• model archive
  • main archive
  • photosynths
• akkerman fortress
  • akkerman fortress project
  • akkerman geophysics
• about us / links
  • about us
  • contact us and links
• the devizes 'wooden heads'
• mobile version

 

4         Recalibrating the literature

Prior to the creation of an English tree-ring chronology in the late 1980s, typology was the main method by which to date a timber structure. Cecil Alec Hewett (1926-1998) pioneered buildings typologies for medieval carpentry joints and timber-framed buildings in south-eastern England (Gibson and Andrews 1998, online). In Hewett’s seminal work English Historic Carpentry the inner sleeve reads “he [Hewett] has shown that the methods of assembling timber buildings, particularly the joints used, follow a strict historical sequence, as datable as ceramics” (Hewett 1980a, inner sleeve). In the case of Hewett, typology is defined as being “historically diagnostic because they are historically unique, that is, they are ‘peculiar to a given time and place’” (Sackett 1977, 371) and therefore, progress from the archaic to the mechanically advanced in a datable sequence of ‘style and function’ (Ibid.).

In 1990, Matthew Johnson warned of relying on typologies to date buildings (Johnson 1990, 247-8) primarily because they are not always reliable and are based on assumptions rather than science. This Chapter will address some previous errors made under the auspices of chrono-typologies forwarded by the likes of; Henri Deneux, Cecil Hewett and J. T. Smith, by applying corrected dates derived through the scientific practice of dendrochronology (Pearson 1997, 30-3; VAG 2000). By ‘recalibrating’ the existing dates arrived at by typologies and informed judgement by Hewett et al, with new solid dates derived from tree-ring analysis, it is hoped Johnson’s warning can be put to one side. This Section will also provide a response to a comment made by Sarah Pearson in which she suggested “one important aspect of construction which is likely to be considerably advanced through tree-ring dating is the typology of timber jointing techniques” (Pearson 1997, 32).

Hewett’s pioneering works have been a starting place for many research projects including this one. Because this thesis aims to test Hewett’s hypothesis that late-medieval timber framed structures can be dated by the joints and carpentry techniques used in their construction in the light of recent dendrochronological advances; it is important that his typology is also informed by tree-ring dating. Hewett himself was unable to achieve this in his lifetime, but this thesis will address by re-visiting his data and updating the chronology based on recent tree-ring data. Once completed (see Table 12) this Section will then retest some of Hewett’s typological assumptions. In doing so, some of his original work will be brought into question and modernised. The resulting recalibrated data will then be compared against the Hampshire data to test for similarities and anomalies.

4.1      Recalibrating Hewett

Hewett’s research focused mainly on the south-eastern county of Essex, though he also carried out various works in other counties – including Hampshire. Before Hewett, very few people would dare ascribe a date for a building to before the 15th or 16th century based on style alone (Gibson and Andrews 1998, online). Hewett changed this by firstly studying the barns at Cressing Temple, Essex, then taking the study further afield and dating other buildings based on the carpentry joints present in them. Using the Cressing barns as a starting point, Hewett would date a building based on whether the joints therein appeared older or newer than those at Cressing Temple. His assumption was that the more advanced a joint or structure appeared, the newer in the sequence it must be –i.e. built after the Cressing barn (Hewett 1962, 240). It can be suggested that Hewett’s methodology was one of teleological progression based on “The hypothesis that carpenters’ joints underwent processes of development towards mechanical efficiency” (Hewett 1980a, 325).

As the table below shows, his technique’s often yielded reasonably accurate results. He could however, be very inaccurate. In several instances he was out by over a hundred years and in one case by as much as 350 years (Table 11).

Table 11 A list of buildings dated by Hewett based on typology and compared against recent tree-ring data

BACK TO TOP

Following on from the above table, it is now possible to recalibrate Hewett’s chronology into one informed by science. The results are shown below, in

Table 12.

Table 12 Hewett's chronology recalibrated by recent tree-ring data and placed in chronological order

If Table 12 is examined, it can be noted that Hewett has two buildings dated to the Anglo-Saxon era. Based on these two buildings - The Church of St. Mary, Sompting, Sussex (Hewett AD950-1050) and the barn at Paul’s Hall, Belchamp St Paul, Essex (Hewett pre-AD 1180) Hewett named Chapter 1 of his Seminal work English Historic Carpentry - “Examples from the Anglo-Saxon Period (AD 449 to 1066)” (Hewett 1980a). Of Sompting, Hewett suggests “the structural method at Sompting is competent, and the workmanship wrought with an assurance that must indicate the previous existence of a long tradition” (Hewett 1980a, 29). Therefore according to Hewett, carpentry was introduced into England in the early Saxon period and reached a ‘competent’ level prior to the Norman Conquest (1066). Both buildings have since been dendro dated to 1300-30 and 1240-75 respectively and therefore, not Saxon carpentry at all. Instead Walker suggests carpentry entered England via the Norman’s around AD1180 (Walker 1999, 28).

Four examples of Hewett’s work will now be analysed in greater detail to illustrate both how accurate, and inaccurate, his dating could be. All four buildings were also surveyed as part of this research in order to gain a greater insight into Hewett’s work and the buildings which formed his research. The four buildings are listed chronologically:

BACK TO TOP

The barley barn, Cressing Temple, Essex

The barley barn is another example of the three different types of dating - typology, radiocarbon dating (14C) - and much later - dendrochronology; all being used to investigate one building.

 

If the dendrochronologically derived dates are taken as the most precise (see Section 2.3) Hewett’s suggestion was very close. Unfortunately, Hewett would never have known how close he was to tree-ring date. Radiocarbon dating, however, was over 200 years out with an unacceptable date range. These three examples, of various dating techniques, highlight the importance of the ability to rely on precise dates, as the implications of inaccurate dates can be profound. Although Hewett was sure his date was accurate, it did lead him to write “this [the radiocarbon date] suggests a date in the eleventh century, centring on 1023, for the felling of the oaks used in its original building. It is perhaps surprising that this date is earlier than the gift of the estate by King Stephen to the Templar’s, but there is no obvious reason why the barn should not have been built whilst the estate was in the possession of the Crown” (Hewett 1962, 271). This would imply that although Hewett doubted the radiocarbon date, he was open to the idea that it could be validated.

Grange Barn, Coggeshall, Essex

Grange Barn was, originally, Radiocarbon dated (14C) to 1130 +/-90 (Essex SMR 8808). Oddly though, Hewett reports the 14C date as being 1020 +/-90 (Hewett 1980a, 47) but later in the same publication he agrees with the 14C date (Hewett 1980a, 289). The associated Savignac Abbey was formed in 1140 and although there is no documentary evidence that gives a construction date for the barn, it was assumed that the building was contemporary (Ibid, 47-8). Hewett had noted the use of open lap joints in its construction (Figure 19) and therefore, based on this joint and its use in the barley barn at Cressing Temple, he suggested the barn was built soon after the formation of the Abbey in the mid 12th century (Andrews 1984). The barn has subsequently been dendro-dated to between 1237 and 1269 (Tyers et al. 1997, 141) and therefore, a late example of an open lap joint (Andrews 1984). Hewett also wrote that “the main posts stood upon stone stylobates about an inch larger all round than the posts’ feet” (Andrews 1984, 49). A stylobate is a raised stone pad upon which the upright post is placed to mitigate rotting. However, the Essex SMR suggests that “observations during restoration suggest the arcade posts were originally set on base plates, as proposed initially, and not on stylobates as suggested by Hewett in 1980” (EssexCC 2003).

 

What should be noted though is that this building, although framed and free-standing does not employ a ground sill for stability. Whereas the Barley Barn at Cressing Temple, Essex (c1200 +/-60) of a slightly earlier date does (Hewett 1980a, 49). The notch lap joints found in this barn, can also be seen at Wherwell ‘stables’ in Hampshire (Figure 20) dendro-dated to 1250 (Roberts 2003, 248). This shows two very similar and coeval joints in separate parts of the country - Essex and Hampshire. This type of joint is also common amongst the majority of the buildings examined by Walker (Walker 1999, 28). Six of the eight buildings used notched lap joints; the implications of which are explained by Walker:

“These dated buildings do not support Cecil Hewett suggestion that there was a development from the late 12th century in the notched lap joint from unrefined entry to secret notched lap. If there was, it was before the late 12th century. Both the unrefined entry and refined entry were being used in the late 12th and early 13th century”. (Ibid.).

The wheat barn, Cressing Temple, Essex

Hewett dated the wheat barn to around 1255 (Hewett 1980a, 102-5). The barn has since been scientifically dated, by dendrochronology, to 1257-80 (Tyers et al. 1997, 51). As Hewett was fairly accurate with both the barns at Cressing Temple, Hewett had a solid datum by which to date other buildings. By his own admission, Hewett would date buildings based on them appearing less advanced, or more advanced than the joints at Cressing (Hewett 1962, 240). As he was so accurate with Cressing Temple barns, one would assume the rest of his chrono-typology would be fairly accurate too. However, the following case study tells a very different story and highlights the need for his work to be recalibrated now dendrochronologically derived dates are available.

St Mary’s Church, Sompting, West Sussex

The timber frame that supports the western tower at St. Mary’s church, Sompting, West Sussex is of a Rhenish helm type (Figure 21). Hewett suggests a date “somewhere between c AD950 and c 1050. It is unlikely to be later than this” (Hewett 1980a, 15). Due to Hewett’s belief that this church roof dated to pre-conquest England, he wrote “The architectural and structural concept of the Rhenish helm is extraordinary, but its execution in carpentry at Sompting is a work of such assurance and competence, achieved with such economy of means, that it both indicates the work of a master and suggests the previous existence of a tradition of framing such works” (Hewett 1989, 15). If this statement were to be true, it would put the introduction of framing back from AD1180 (Walker 1999, 28) to AD950, before the introduction of the French style, even before the Norman Conquest (Hewett 1982, 341). Clearly then it is essential to the understanding of the evolution of carpentry, that this building be scientifically dated, in order to validate whether 950 or 1180 are to be used as the start of the carpentry tradition in England. Indeed, Hewett assumes “the structural method at Sompting is competent, and the workmanship wrought with an assurance that must indicate the previous existence of a long tradition” (Hewett 1980a, 29). Therefore, Hewett is suggesting carpentry existed in England before AD 950 based on Sompting alone suggesting “it is no longer possible to ascribe the introduction of any types [of carpenters’ joints] to the Conquest, or the Normans” (Ibid.).

Hewett also notes the use of carpenters’ marks in the form of chisel cut Roman numerals, suggesting that the frame was measured and framed elsewhere. Therefore, Hewett suggests the carpenters who constructed the frame at Sompting had “anticipated ensuing carpenters’ methods” which would not be seen again until 1180 (Hewett 1989, 15). Even though Hewett was convinced the roof was Saxon in origin, he noted that some of the joints – “a tenon with one shoulder ‘scribed’ to fit over a waney edge” – are unknown elsewhere until the 13th century, yet he remained convinced he was looking at a Saxon roof.

Fortunately, Hewett’s date was off by around 380 years (Aldsworth and Harris 1988, 140; Pearson 1997, 33). In 1985 two timber samples were Radiocarbon dated to AD 1323 +/-51 by Jill Walker. A tighter date range was sought and in 1987, 17 samples were taken for dendrochronological analysis giving a new date range of AD 1300 to 1330, by Ian Tyers (Aldsworth and Harris 1988, 140-3; Tyers 1990, 45). It should be noted that the main body of the church is of a pre-conquest date, it was added to in the later part of the 11th century which, it seems, Hewett based his date on (Aldsworth and Harris 1988, 139). Before the recalibration of Hewett’s dates for St Mary’s church, it is plain to see the implications that an incorrect date can have our knowledge of the past. 380 years in the medieval to late medieval period sees many profound historical changes - from Saxon England, through Norman occupation, to the beginning of the 14th century dominated by famine, plagues and war.

 

BACK TO TOP

4.1.2 Summary

Hewett has published many articles and books, mainly on the subject of joints and structural carpentry. Any student of structural carpentry is therefore, a student of Hewett’s work. Thus, it is important that his dates and assumptions are re-examined in the light of dendrochronological data when, and if it becomes available. This needs to be done to forward the discipline, not prove Hewett right or wrong. Hewett worked with the data available to him and when Radiocarbon dates were available, he would use them, or at least acknowledge them, because as we have seen the 14C data is often more inaccurate than Hewett ever was. This Section has highlighted how important the recalibration of his chrono-typology is. When fundamental principles of carpentry are pinned to an inaccurate timeline and alternative method of dating must be sought. For now, that method is dendrochronology. With it, previous dates, such as those given by Hewett can be revisited and updated in the light of recent advances in the field.

This Section has illustrated several inaccuracies in Hewett’s method. The two structures he suggested provide evidence for an Anglo-Saxon carpentry tradition were misinformed by inaccurate dates. Instead, it would appear that carpentry developed in England around 1180 (Walker 1999, 28) and not prior to 950 as Hewett suggested (Hewett 1980a, 29). The following Section will also prove Hewett’s evidence for a transitional scarf c1350 from splayed to halved scarf is also misinformed. It must be noted though, that Hewett was right about a great deal of other important facts relating to joints. Much of which shall be evidenced in the following Sections regarding joint typologies and chronologies.

 

BACK TO TOP

 

 

 

 

 

 

Friday, March 18, 2011 March 18, 2011 in the UK | Copyright 2007 by Richard John Haddlesey PhD research student at the University of Winchester | Contact