In the last chapter I outlined a general history of lithic research in Ireland, describing some of the main preoccupations of researchers. In this chapter I will turn attention specifically to quartz research in Ireland. Section 3.2 discusses the position of quartz research here; I will discuss the neglect of quartz research, and how this has had implications for misunderstandings of the prehistoric use of quartz. I will focus on two areas that have produced concentrations of quartz and have recently had parts of the assemblages reassessed – Lough Gur, Co. Limerick and sites from north Co. Mayo. This will be followed in Section 3.3 by a discussion of the findings of the database of quartz finds. This database has been formulated by a literature review, the excavations.ie database, and an archive search in the National Museum, Dublin, and has highlighted the widespread use of quartz throughout Ireland.
As we saw in Chapter 2 quartz had been identified in the archaeological record from the time of Knowles’ work in the late nineteenth century; the amount of quartz collected was minor, however, and made no impact on general discussions of prehistory. In Macalister’s (1949, 179) second edition to The Archaeology of Ireland, quartz is cited in relation to magical stones, such as “so frequently found in tomb-deposits”. Macalister’s publication came out just as Ó Ríordáin (Ó Ríordáin and Ó Danachair 1947) began publishing the Lough Gur (Figure 3-1) excavations mentioned in Chapter 2, which were the first excavations to produce a lithic assemblage with a substantial quartz component.
Figure 3-1 Locations mentioned in Chapter 3
In Ó Ríordáin’s articles, however, little discussion was undertaken on the lithic assemblage, with more emphasis placed on the pottery assemblage as well as on the structural components excavated. Subsequently, while quartz tools have continued to be found over the years, they have rarely been mentioned in the literature, and prehistoric quartz has primarily been viewed in the context of megalithic monuments, especially as these types of sites were chosen for exploration and excavation, thus compounding this view. A prime example of this can be seen in Waddell’s (1998) introductory textbook on Irish prehistory – while quartz is mentioned in the text, for instance in relation to one of Irish archaeology’s most famous icons, the quartz facade at Newgrange, the index has just one entry for quartz, and ironically this is related to quartz found during the excavations which uncovered a series of pits, stake holes and ditch with Neolithic artefacts at Langford Lodge, Co. Antrim (Figure 3-1), the premier flint region of Ireland.
In Woodman et al.’s (2006) recent overview of quartz use in prehistoric stonecraft, they noted eight sites with quartz and mentioned that it occurs at a number of other sites. They suggest that “[a]lthough quartz is relatively common across Ireland as vein deposits, it is often poor grained and has a sugary texture, making it unsuitable for flaking”, and, comment on the difficulties faced in the identification and classification of quartz artefacts (Woodman et al. 2006, 84). They make the distinction between vein quartz and rock crystal, suggesting the latter was used more frequently for stone tools.
The fact that the widespread use of quartz as stone tools has not been acknowledged can lead to misconceptions within the archaeological record. For example, the folklorist Thompson (2004), calling for a greater degree of openness on the part of archaeologists towards the insights of the sì tradition in folklore in their research, uses the example of quartz use in Ireland in prehistoric and historic times. Citing a paper by Gibbons and Higgins, Thompson (2004, 359) notes that they regard the use of quartz in monumental structures as having elusive ritual meanings; he argues, however, that this is incorrect and that quartz is mentioned in folklore and much can be learned from these – primarily, quartz is not to be used by the living but reserved for places of the dead, and quartz is intimately related to the sì (the sì meaning both the burial mounds and the spirits of the mounds). He goes on to argue:
"[t]his reveals continuity at several levels: one is simply the remembrance of Newgrange as “shimmering” and of its use of white quartz crystals...Even more pronounced, if more subtle, is the common thread that white quartz crystals (clocha geala/clocha uaisle), are closely associated with the sì, and are appropriate only for thanatopic architecture. This is true whether in the ancient sites, old church graveyards, or in the graveyards and folk beliefs of traditional Ireland today. There is no effective break in this tradition, from the Neolithic on down, although it may be currently less pervasive than in the past" (Thompson 2004, 360-1, emphasis added).
Therefore, we can see that as a consequence of archaeologists failing to bring attention to the widespread ‘mundane’ use of quartz, researchers from outside the field have a skewed sense of the role that quartz played in prehistory; while quartz was clearly a substance that held special ritual properties and purposes, it was also used by the living in their stone tool repertoires, and therefore there would appear to be a break in the tradition of quartz use since the Neolithic and Bronze Age.Top of Page
The Lough Gur (Figure 3-1 and Figure 3-2) excavations were the first to reveal a substantial quartz component within a lithic assemblage; the intensity of excavations in this area, as well as the building of a gas pipeline nearby, led it to have the greatest concentration of quartz finds in Ireland. Quartz was found at nearly all of the prehistoric sites excavated (Ó Ríordáin and Ó Danachair 1947; Ó Ríordáin 1951, 1954; Ó Ríordáin and Ó h-Iceadha 1955; Grogan and Eogan 1987; Woodman and Scannell 1993), as well as a minor amount from Rathjordan (Ó Ríordáin 1947) and Cush (Ó Ríordáin 1940) which were also excavated at the time. Added to these are the more recent finds from the gas pipeline excavations (Gowan 1988) – some of which can be noted as the relatively straight line of finds in Figure 3-2 – and another research excavation at Chancellorsland, Tipperary (Doody 1993b).
Figure 3-2 Lough Gur area's sites with quartz. Inset: Lough Gur sites with quartz
Ó Ríordáin’s (1949; 1951; 1954) excavations of various Neolithic and Bronze Age settlement and ceremonial sites at Lough Gur began in the 1930s and continued for 18 years; his excavations were added to in the 1950s by Liversage (1958), and in the 1990’s by Cleary (2003). Thousands of lithics were collected during the various excavations (Woodman and Scannell 1993), along with over 400 axes (2006a) – however, it is unclear what collection strategy was used during the excavations and what the retention policy was during post-excavation. In Woodman and Scannell’s (1993, 54) table of the raw materials from a selection of six sites, there are 1300 flint, 219 quartz, 188 chert, and 87 greenstone artefacts, showing a predominance of flint; however, this selection is skewed as the chert-dominated site – Site 10 – is not included. (The quartz database in this thesis provides a minimum count of 457 quartz artefacts for all of the Lough Gur sites, see Section 3.3.)
There is a debate as to why chert may be under-represented in the record, in a region where chert would be available – one possibility given is that Ó Ríordáin may have initially overlooked chert, and that it occurs more often during his later excavations when he became more familiar with the material (Woodman and Scannell 1993, 54); the authors suggested that the Lough Gur assemblage probably represents “an accumulation through time of a limited selection of highly curated artefacts, particularly arrowheads and axeheads” and that this “may have to do with specialist activities and ritual as with settlement”. Woodman and Scannell (1993, 54) suggested that the predominance of flint follows a similar pattern elsewhere in the southwest for later prehistory, even in areas where chert was available, and that the use of quartz in the area “may be a simple expedient use of a convenient resource”. This comment is similar to Herity’s (1987, 146) suggestion that chert was used as a substitute for flint, as opposed to it being a valid raw material in its own right.
Woodman and Scannell’s (1993) reassessment of part of the Lough Gur lithic assemblage was based on Scannell’s (1992) thesis which looked at the effect of raw material availability on the prehistoric lithic traditions in the area. She noted that some of the excavated material was unobtainable, including the quartz from Site A. Another problem encountered was that due “to the nature of some of the quartz in the assemblages, coarse grained and brittle, it was difficult to identify pieces which had been struck and in some cases pieces which had been retouched” (Scannell 1992, 22). While the Lough Gur assemblage consisted of both vein quartz and rock crystal, with many of the sites dominated by rock crystal (see Grogan and Eogan 1987), Scannell’s (1992, 30) metrical analysis does not differentiate between them, and she describes the rock crystal as “fine grained” which suggests fine grained vein quartz instead of actual crystals. Therefore it is unclear whether these artefacts are indeed vein quartz or crystal. Scannell’s analysis does not appear to take the fragmentation of the quartz into account and the various metrics for the flakes/blades provided do not appear to differentiate between complete and fragmented flakes/blades. Scannell described the use of quartz at Lough Gur as a “necessity” (1992, 38), and argued that “[p]ercussion hammering seldom produces the required tools, However [sic] bipolar percussion offers a more advantageous method of producing the preferred tool types”. Although no references are cited, this would seem to follow the common notion of the ubiquity, indeed the necessity, of a bipolar technique when working quartz (see Chapter 4). Furthermore, while she noted an apparent preference to use the bipolar technique to knap quartz, she does not consider that the bipolar cores may have been first reduced as platform cores, as implied by the amount of non-bipolar quartz flakes and blades in the assemblages.
Two Neolithic houses were excavated at Tankardstown South close to Lough Gur (Figure 3-2) and produced a small lithic assemblage of 72 artefacts, predominantly flint; Woodman (n.d.) noted that the assemblage bore a resemblance to Sites A and B at Lough Gur, and that the diagnostic types (leaf-shaped arrowhead, scrapers, and retouched pieces) fitted with Early Neolithic type lithics. The quartz consisted of a possible platform core and a possible scalar(bipolar) core, three rock crystal flakes, along with pieces deemed to be natural: one pebble, three fractured pebbles, three small fragments, and one crystal (Woodman n.d.). While no spatial contexts were provided, Smyth (2006b) notes that the quartz material came from both houses.Top of Page
The north coast of Mayo (Figure 3-1 and Figure 3-3) has seen an intensity of archaeological activity instigated by the discovery of prehistoric field boundaries under the bog; parts of these have been probed, primarily around the Céide field system, along with a number of excavations of the boundaries and related settlement sites (Caulfield 1983), which for the most part have not been published. The findings of the field systems led to the excavation of megaliths at Behy and Ballyglass; the latter excavations consequently uncovered house and hut sites (Ó Nualláin 1972, 1998), and also to development-led excavations in preparation for the interpretive centre at Céide (Byrne 1989b), as well as more the recent research excavations at Rathlackan (Byrne 1993b) and the lithic scatter at Belderrig (Warren 2009b), the primary case study of this thesis. These excavations have almost invariably produced quartz finds, ranging from a minor percentage of the assemblage to a majority of the assemblage. With the excavations having taken place under different standards of recording and indeed different concepts of artefact retrieval and retention – even for ‘standard’ material such as flint – it is difficult to provide a clear assessment of the raw material variability at the sites, especially in terms of the quartz component.
Figure 3-3 North Mayo. Land use and archaeological sites. Land use adapted from Corine dataset (EPA 2004a)
The assemblages from the Behy and Ballyglass no. 13 court tombs, originally excavated in the 1960s and 1970s, have recently been analysed for the forthcoming publications (Dolan and Warren 2006; Warren Forthcoming). A second court tomb at Ballyglass no. 14 and associated structures was excavated at the same time, belatedly published in 1998 (Ó Nualláin 1998). A fourth court tomb, at Rathlackan, was excavated in the 1980s, and has not been published (Byrne 1993b).
Excavations initially took place at Behy, and Dolan and Warren (2006) noted that the “notebooks for Behy suggest that very early on in the first season of excavation the importance of quartz was recognised and the difficulties in identifying bulbs of percussion on the material were noted”; just under 63% of the lithic assemblage was quartz (total assemblage 597), along with two rock crystal flakes/blades. The non-quartz assemblage was primarily non-retouched flakes of chert and flint, with two flint bipolar cores and a chert platform core; the retouched component consisted of over a third concave scrapers, along with other scraper types, projectiles, and various other types (Dolan and Warren 2006). Six platform and five bipolar quartz cores were noted, with some of the bipolar cores being bipolar-on-platform cores; four of the quartz artefacts were retouched – two as points and two as scrapers (Dolan and Warren 2006). Warren and Dolan (2006) noted that 57% of the quartz artefacts was broken; looking at the identified flakes only, 71% (n=190) was fragments. Warren and Dolan (2006) commented that “is not surprising that the quartz has a high incidence of breakage as it is prone to fracture as part of the knapping process”. This rate of breakage is, however, actually much lower than that noted during the present analyses of the experimental knapping and two case study assemblages – 88% of the experimental flakes was fragments as was about 95% of the two case study assemblages’ flakes (see Chapters’ 6, 9 and 10). This may suggest that there was a greater emphasis on the deposition of complete quartz flakes at Behy, or may suggest that the excavators retained more complete flakes than fragments.
At Ballyglass no. 13, which was excavated shortly after Behy, quartz forms less than 9% of assemblage (total assemblage 805), and Warren (Forthcoming) suggests that the retention during excavation and post-excavation appears to have been low and should be seen as a minimum count. Ballyglass no. 13 uncovered a court tomb, which partially overlay a rectangular house, and had a series of pits outside the tomb. The non-quartz component was primarily chert with about 20% flint and dominated by non-retouched flakes; 25% of the entire assemblage was retouched and 45% of this was concave scrapers, with the rest consisting of other scrapers, projectiles, and various other types (Warren Forthcoming). The assemblage was suggested to represent a Middle to Later Neolithic assemblage (Warren Forthcoming). The quartz (62 quartz and 9 rock crystal) was dominated by non-retouched flakes, with five bipolar cores (four rock crystal). Most of the quartz came from the pits, with eight artefacts from the house and one non-worked rock crystal from the court tomb. One rock crystal leaf-shaped arrowhead came from a posthole in the house (Warren Forthcoming). This appears to be the only example of a quartz projectile in the Irish archaeological record.
As with the quartz from Ballyglass no. 13, at Ballyglass no. 14 the analysis of the quartz may be problematical – while a substantial amount of small fragments of quartz (2.5 kg) were collected from the topsoil during excavations, only two were regarded as “unequivocal artefacts”, out of a total assemblage of 2300, excluding the discounted quartz (Ó Nualláin 1998, 140). At Rathlackan, the initial reports stated that quartz comprised 23% of the lithic assemblage (total assemblage 700) (Byrne 1993b).
About 50km south as the crow flies from the Ballyglass house, a Neolithic house was excavated at Gortaroe, west Co. Mayo (Figure 3-1) with dates of 3623-3105 cal BC from the foundation trench and 3907-3641 cal BC from an internal posthole (Gillespie Forthcoming). Here, 261 lithic artefacts, of which 86 were quartz; the type of quartz was not mentioned, which may suggest that it was vein quartz with no rock crystal (Milliken Forthcoming). The chert and flint assemblage was primarily flakes, with a javelin head, a number of arrowheads and scrapers; the four arrowheads were leaf-shaped arrowheads while nearly half (n=4) of the scrapers were hollow scrapers (Milliken Forthcoming). Leaf-shaped arrowheads are seen as earlier Neolithic types, while hollow scrapers are seen as later Neolithic types (Woodman et al. 2006). The non-quartz assemblage was seen as primarily bipolar-based with all the cores (n=6) identified as bipolar; the quartz artefacts (57 “flakes”, four “flake fragments”, one “fragment”, and 24 “chunks”), however, remained unassigned to a technique (i.e. either bipolar or direct percussion). The quartz and non-quartz artefacts came from the foundation trenches (17 quartz), the house interior (one quartz), internal pits (70 quartz) and external pits (two quartz). The flake breakage rate of 6.6% (4.6% of total debitage) is extremely low; as noted in relation to Behy above, 88% of the experimental flakes of this project were fragments.
Back to the north Mayo coast, other quartz finds from the area are from the pre-bog field system and Bronze Age settlement at Beldergbeg (unpublished (Caulfield 1971b)), and the lithic scatters identified at Belderrig, Céide fields, and Glenulra. The latter lithic scatter was identified after an extension to a road built through the bog unearthed lithics which were subsequently spotted by chance in the spoilheap, and led to the excavation of the disturbed area as well as the trowelling of most of the initial spoilheap; the assemblage amounted to 1100 lithics, of which 25% were quartz (Byrne 1992a). These assemblages have not been analysed.
Figure 3-3 shows the four excavated court tombs along with the distribution of the other court tombs and possible early-type megaliths on the north coast of Mayo. This map also shows the relationship between excavated finds and non-excavated finds from the area. One of the key issues in assessing distribution maps of lithics concerns the tradition of collecting in a given area – in general, there is no tradition of collecting in the west of Ireland, inevitably entailing less finds. Although chert is the most common raw material found during excavations in the west, flint is invariably more common as a non-excavated find, along with stone axes, no doubt due to the latter’s size and distinctive shape (see Driscoll 2006). Apart from the lithic scatter at Glenulra already mentioned, there are nine findspots of lithics in the area shown in Figure 3-3. Following the general pattern for the west, five of these are finds of single stone axes, two are single flint arrowheads, and one collection from a townland consisted of a flint knife, two chert scrapers, an awl, and a hammerstone collected during tillage; the others were from both the bog and tillage. The ninth findspot is at Belderrig, where Seamas Caulfield noted the quartz scatter being eroded out of the cliff face. Hence, while quartz is prevalent in the excavated assemblages the only non-excavated quartz collected was by an archaeologist. Added to this problem of quartz identification is that much of this region is covered in blanket bog, making any lithic scatter surveys extremely limited: that being said, as a result of Warren’s (2009b) excavation at Belderrig, and having archaeologists conscious of quartz in the area, a number of other quartz scatters near Belderrig have now been identified.Top of Page
At the initial stage of formulating this research programme, a survey of the literature and the online database excavations.ie, highlighted that quartz had been recorded in various contexts throughout the country. In order to bring this to the attention of the archaeological community, it was decided to set up a database of all known quartz finds from prehistoric contexts in Ireland (i.e. excluding the widespread medieval use of quartz). Where it forms part or all of the architecture of monuments, quartz has been excluded, though it is acknowledged that a division between the artefactual and architectural aspects of quartz is in a sense arbitrary – the deposition of a quartz core at the base of a post in the Neolithic house at Cloghers, Co. Kerry (Kiely 2003, 184) (Figure 3-7) could be interpreted as a deposited ‘find’ or as an integral part of the original ‘architecture’ of the structure.
This database has been formulated by a literature review including grey literature, the excavations.ie database as mentioned, and an archive search in the National Museum, Dublin. Along with the 12 maps presented below, the database can be queried using the Arcview interactive map on the accompanying CD-ROM, as well as viewed in the raw data in the accompanying Excel file on the CD-ROM. While the full database is presented on the accompanying CD-ROM, Appendix A-2 presents a summary account of the worked quartz finds and Appendix A-3 presents a summary account of the non-worked quartz finds.
It is important to note that this database is constrained by the primary data in five key areas:
- Firstly, the database includes both worked and non-worked finds – in other words, deposits of quartz in ritual and funerary contexts that do not necessarily include ‘worked’ quartz (Herity 1987; O'Brien 1999). ‘Worked’ ‘possibly worked’, and ‘non-worked’ finds are differentiated in the database, and all but one of the maps presented here exclude ‘non-worked’ finds. This distinction between worked and non-worked, however, is based on the description in the primary source, and, as Warren and Neighbour (2004) have argued, the differing contexts of quartz use are associated with different sets of archaeological terminology. As well, there are significant difficulties in the identification of worked quartz even by experienced lithic analysts (see Chapter 7). Many ‘non-worked’ finds may, actually, include worked material; and many ‘worked’ finds may need considering in the light of the use of quartz in ritual contexts.
- Secondly, related to the above constraint, the database has not checked the identification of the quartz as worked or possibly worked, but rather simply cited it as stated in the reference.
- Thirdly, this count should be seen as a minimum amount, as it is apparent that even though quartz may have been found during excavations, it may not be stated explicitly as quartz in the reports or publications but instead called ‘stone’ artefacts/lithics; this also applies to the National Museum archives, where quartz is currently listed as stone, with only flint given a separate category (Cahill pers. comm.).
- Fourthly, the point made earlier, that quartz will be under-accounted for in both surface collections and excavations, must also be borne in mind.
- Finally, due to the nature of the primary sources used, it has not been possible to show the ratio of quartz to non-quartz lithics, as amounts are often not given, or are difficult to interpret. It has also not been possible to provide a clear breakdown in the ratio of vein quartz to rock crystal artefacts.
These problems demonstrate that any current ‘total’ for quartz in Ireland would be incorrect – but in a context where over two thousand licences for archaeological excavation are granted each year in Ireland in recent years (Anon. 2006a, 8), any static figure would be meaningless in any case. More significantly, and accepting the caveats above, this rudimentary database highlights the extent of the quartz ‘problem’ in existing archives. As mentioned, Woodman et al.’s (2006) recent review of Irish prehistoric stonecraft highlighted eight instances of quartz lithics and mentions that quartz occurs at a number of other Neolithic structures. This database has shown that at least 170 townlands have quartz artefacts that are described in the literature as either worked or possibly worked, with many of these townlands having more than one findspot/structure/feature with quartz.
Figure 3-4 presents the quantities of quartz found – while this suggests that the numbers are generally low with most being single finds or under 10 artefacts, this has to be taken in context of the total lithic assemblage collected. For instance, while only five quartz lithics were found during excavations of the Neolithic structure at Drummenny Lower, Co. Donegal, these accounted for 33% of the total lithic assemblage (Dunne 2003), and at the Neolithic structure at Enagh, Co. Londonderry the quartz finds were a “few” flakes with only one flint artefact recovered from this partial excavation (McSparron 2003). Conversely, Figure 3-4 shows a large assemblage of quartz from the Neolithic site of Donegore Hill, Co. Antrim (Figure 3-1) – a total of 341 artefacts – but this only represents 1.4% of the total assemblage, the rest being flint (Nelis 2003, 207). It also shows a large assemblage (272 artefacts) from the Neolithic house site at Ballygalley, Co. Antrim (Figure 3-1), but these were found along with hundreds of thousands of non-quartz lithics (Jeremy et al. 2002). The right map in Figure 3-4 provides a grouped total for certain sites such as the Thornhill Neolithic enclosure whose numerous structures and pits together totalled over 800 quartz artefacts, and the four Neolithic houses at Corbally which together produced over 50.
Figure 3-5 shows the decade in which the quartz finds were made. The chronology of the finds highlights two main aspects: the minor amount of quartz noted before the second half of the twentieth century, and the significant increase of finds in the 1990’s and the 2000s up to 2006 – this stems mainly from the significant increase in development-led excavations in the last decade, and can be seen as especially concentrated on the east coast (Figure 3-6). These maps should also be seen in the context of the general distribution of finds of other raw materials; for instance, the southwest and the midlands both have a relative lack of non-excavated finds of any material, and both have experienced less research and fewer development-led excavations. Figure 3-5 also highlights that where a time period was noted, the clear majority are placed in the Neolithic.
The top left map in Figure 3-6 presents the find circumstance of the quartz with a distinction between surface finds found by ‘amateur’ collectors or the general public and fieldwalking finds found during research-led projects. The majority of the findspots are from excavations. Figure 3-6 (top right) shows whether the quartz was collected during either research or non-research based activities – non-research based activities include development-led excavations as well as finds by ‘amateur’ collectors or the general public, which account for a small amount of the total finds. Given that research excavations comprise a small proportion of the overall excavations in Ireland, the high proportion of the quartz from research excavations is noteworthy.
Figure 3-6 (bottom left) shows the context of the quartz found during research projects. The two groups of finds in the southeast from ploughed fields were collected during the Bally Lough Project (see above pp. 24-5); the amount of finds collected was small, usually one per field surveyed – the area in between these two groups was not surveyed, creating an apparent gap in distribution. The group of ploughed field finds on the east coast are from the Mt. Oriel and Boyne Valley surveys (see above p. 24). Again, the amount collected was small, but they highlight that where research projects have dedicated time to ploughzone surveys, they have turned up evidence for quartz. In this context, Kimball’s comments cited in Chapter 2 (see above p. 26), on the probable under-representation of quartz in his project’s survey in Donegal may well apply also to other surveys (see also Bradley 1995). (Kimball’s survey’s quartz finds are the two ploughed field findspots in the north.)
Figure 3-6 (bottom right), showing the find contexts from other than research projects, highlights the lack of quartz found during either digging or surface finds by the general public – as mentioned, this contrasts markedly with the amount of finds of flint or stone axes from such contexts around the country. This lack of quartz finds by the general public arises from both the difficulty in identifying quartz artefacts and the lack of expectation that quartz artefacts might occur because so much of the story told of prehistory in Ireland is based around flint. As is clear, excavations and monitoring account for nearly all of the non-research finds. Figure 3-6 (top left), showing the find circumstance of both research and non-research together, highlights this trend.
Figure 3-7 provides a breakdown of the site types where quartz has been found, excluding fieldwalking, surface, and non-monument digging contexts; it includes one map (bottom right) showing sites with only non-worked quartz noted. The top left map shows the finds in megaliths, stone circles, and cairns. The large amount of court tombs with worked quartz highlights how many of these megalith types have witnessed research excavations, but may also suggest a greater emphasis towards quartz at court tombs compared to portal or wedge tombs; only one portal tomb appears to have worked quartz and another non-worked quartz, and only one wedge tomb appears to have worked quartz and two with non-worked quartz. Worked and non-worked quartz is also frequently found at passage tomb sites, including the Passage tomb complex at Carrowmore, Co. Sligo whose monuments are listed as ‘boulder circles’ in the top left map in Figure 3-7.
The top right in Figure 3-7 presents finds from pits, cists, barrows, mounds, middens, a causewayed enclosure, houses, and structures/settlements; the Neolithic houses and possible houses are named on the map. 14 Neolithic house groups have worked quartz finds with quartz found at more than one house at Thornhill (three houses), Ballygalley (two houses), Lough Gur (seven houses) Corbally (four houses), and Tankardstown South (two houses); Lough Gur also has a number of Bronze Age houses with quartz. Therefore, at least 27 Neolithic houses contain quartz lithics. Apart from the seven from Lough Gur, the rest are all Early Neolithic examples (see Smyth 2006b). The quartz from these Neolithic houses very often came from internal and external foundation trenches, internal features such as pits, external pits associated with the structures, and, in rare cases, occupation layers. The bottom right map in Figure 3-7 highlights that three Neolithic houses with no worked quartz contained non-worked quartz; at Russellstown, “unworked” quartz was found in the foundation trench (Logan 2007), at Kishoge a crystal was found in the topsoil (Milliken n.d.), and at Corbally House 5 an “unworked” crystal fragment was found in the topsoil (Milliken 2002b).
Lough Gur, Ballygalley, and Corbally had significant proportions of rock crystal. Ballygalley’s quartz assemblage was 48% rock crystal (Moore 2002). While 29 vein quartz cores and four rock crystal cores were noted, none were designated as bipolar even though the raw material appears to have been pebbles and small cobbles and the cores appear to be extremely small (Moore 2002), much smaller than the platform cores from Thornhill (see Section 10.4.3). Along with a couple of scrapers and a few retouched flakes, 10 non-retouched crystals had evidence for use with Moore (2002) suggesting that “perhaps the serpentinite beads were bored with the rock crystal prisms”. Moore (2002) also points out that quartz was used as a temper for the pottery at Ballygalley, where it was the sole temper used, “particularly for the Carrowkeel type vessels…[which are]…generally associated with funerary activity”. As noted by Cooney (2000), other white inclusions in the form of bone and shell temper are also noted in Carrowkeel pottery. Moore (2002) suggests that the “quartz and rock crystal occur only in the Middle and Later Neolithic phases” at Ballygalley, and he argues for a similar chronological pattern of quartz use at Lough Gur. However, this chronological use of quartz does not appear to apply to the Early Neolithic houses at Cloghers, Tankardstown, Corbally, or Enagh (for lists of C14 dates for these houses see Smyth 2006c, 239).
The excavations at Corbally uncovered seven Neolithic houses with quartz from four. Houses 1 to 3 were excavated first: the three houses produced 299 lithics of which 64 were quartz and nearly half were rock crystal; one possible scalar (bipolar) core and one possible core were noted (Brady 2001) (Table 3-1). The non-quartz component – primarily flint – consisted of bipolar and platform cores, flakes, and a series of leaf shaped arrowheads, scrapers and various others, as well as a number of stone axes (Brady 2001). The majority of lithics came from the foundation trenches of the three houses (Purcell 2002) For the quartz, it was suggested that 17 of the flakes were “utilised” but the term was not defined; a number of flakes were posited as being either knapped with “direct soft hammer percussion” or “indirect soft hammer percussion” but how this was determined was not defined (Brady 2001). The fourth house at Corbally produced 115 lithics (four bipolar cores and no platform cores) – predominantly flint – of which two were vein quartz flakes and one was a rock crystal flake; there were also 17 “unworked fragments” (Milliken 2002b). Milliken (2002b) noted that one of the flakes had no platform, and in discussing the attributes for the differing techniques of hard and soft hammer that “hard hammer…produces flakes with platforms…[and] soft hammer…sometimes produces flakes with no platforms”. A lack of platforms is not usually an attribute cited for signifying soft hammer percussion (this is discussed further in Section 5.6 and Chapter 6).
|House 1||House 2||House 3|
|Vein||Rock crystal||Vein||Rock crystal||Vein||Rock crystal|
|Possible scalar core||1||-||-||-||-||-|
Table 3-1 Quartz finds from three Neolithic houses at Corbally (compiled from Brady 2001)
Figure 3-4 Worked quartz finds. Left: quantities. Right: quantities for sites grouped
Figure 3-5 Worked quartz finds. Left: find decade. Right: find dating
Figure 3-6 Worked quartz finds. Top left: finds circumstance. Top right: Find context. Bottom left: Research find circumstance. Bottom right: Non-research find circumstance
Figure 3-7 Quartz finds from selected site types, excluding surface finds. Bottom right: non-worked quartzTop of Page
The review of Irish quartz research has shown that the amount of work has been minimal, with no intensive research carried out. Whereas experimental work has been conducted elsewhere to understand the fracture mechanics of quartz (see Chapter 4), such experimental work has not been carried out in Ireland, and very few references are made to this experimental work in Irish analyses. As suggested in reference to the article by Thompson (2004), this minimal amount of research has led to erroneous understandings of the character, and extent, of quartz use in Irish prehistory. Quartz has been perceived as a material primarily related to megaliths, and seen in terms of architecture or ritual depositions and funerary contexts; where quartz has been noted, such as at Lough Gur, it has been seen as a poor substitute for other materials – a raw material for expedient use, and to be used out of necessity rather than choice. For the most part quartz assemblages recovered to date have been small in number, and the larger assemblages that were collected have not been subject to careful scrutiny. The examples from North Mayo and Lough Gur have shown that quartz has posed problems for excavators in terms of the initial identification of quartz during excavation, as well as in post-excavation analysis.
The database highlights that quartz lithics have been discovered from all over Ireland in a wide range of archaeological excavations. In terms of quartz from non-excavated contexts, quartz is more than likely under-represented in finds by archaeological researchers as well as by ‘amateur’ collectors and the general public compared to flint artefacts. With the increasing amounts of development-led excavations, more quartz has been found, but it is difficult to assess the level of awareness of quartz identification in Irish archaeological circles. This difficulty applies equally to research excavations. In a British context Warren and Neighbour (2004, 84) have suggested that “[i]n the field it is still common to see worked quartz missed by even skilled and experienced excavators”. It is probable that this applies to Ireland as well. Moreover, the quartz recognition experiment discussed later in Chapter 7 highlights the significant difficulties that even experienced lithic analysts have in identifying and classifying vein quartz artefacts. Therefore, it is difficult to get a clear picture of the prehistoric use of quartz from the various assemblages outlined here.
As well as the quartz from megaliths, especially court and passage tombs, numerous barrows, pits, and cists have quartz deposits. For the Neolithic houses, quartz lithics have been found in 27, and 20 of these were Early Neolithic, accounting for about a quarter of the known early houses. As well as these worked quartz finds, another three houses contained non-worked quartz. The quartz from these Neolithic houses very often came from internal and external foundation trenches, internal features such as pits, external pits associated with the structures, and, in rare cases, occupation layers. Where cores have been noted, they are generally bipolar (or called scalar) with Scannell, for example, suggesting the necessity of using a bipolar technique with quartz; conversely, at Ballygalley 33 cores were noted, but none were described as bipolar even though many appear to be very small cores.
Overall, most analyses do not appear to have taken the fracture characteristics of the quartz into account, and this may partly explain the differences in core types and other types noted (see Chapter 7); others that have been cognisant of the different fracture mechanics of the quartz appear to have underestimated the fragmentation rate, and possibly the fracture characteristics, of the material. A clear pattern for all the quartz assemblages is the minor proportion of formal ‘types’ such as scrapers and projectiles compared to other materials in the same assemblages, with the most commonly assigned type being scrapers. While this may relate to the under-recognition of retouch on quartz (see Lindgren 1998, see also Chapter 7 for the over-recognition of retouch), it is also possible that it relates to a different use of quartz – of use in a non-retouched form – that does not ‘fit’ easily into the archaeological record, and archaeological understandings, from a typological perspective. This important point is discussed further in the next chapter which outlines quartz research that has been undertaken outside of Ireland.
The term ‘quartz’ is generally used in the literature to cover both vein quartz (xenomorphic quartz) which consists of an aggregation of quartz crystals and individual quartz crystals (automorphic quartz) usually called rock crystal. Therefore, in the following literature review ‘quartz’ by itself is generally a shorthand for vein quartz, but also can mean both xenomorphic and automorphic quartzes.[return].
A number of researchers use the term 'scalar core' to mean cores produced using a bipolar technique.[return].