Site Section Links
KRONOS Vol VIII, No. 3
PETROFABRIC ANALYSIS: AN UNRELIABLE ARCHAEOLOGICAL TOOL
It is not uncommon for archaeology to borrow the techniques of other disciplines.
Thus the employment of the soil scientist, J. S. Bibby, by the archaeologist, E. W. MacKie,
follows standard archaeological procedure. The subsequent claims that Bibby's analysis showed
the Kintraw stone layer to be man made,
A survey of the relevant literature indicates that petrofabric analysis has hitherto been mainly used in the investigation of glacial and aquatic movement. Work has also been performed on solifluction deposits and talus formation; and, as little work has been done in the corroboration of the findings in these two fields, we should firstly investigate the use of this technique in its more normal applications.
In 1932, K. Richter
Petrofabric techniques are also utilized to demonstrate the direction of glacier flow. Blanket acceptance of the results obtained by this method, however, can lead to considerable error in the correct interpretation of the direction of glacier flow. In a number of cases, the long axes have been found to lie normal to the direction of the ice flow. As a precaution, geologists have found it necessary to determine the path of the glacier from other geological evidence(s). Concordances between conclusions obtained from striae and from petrofabric considerations demonstrate, with high confidence, the true direction of glacier movement. If, however, the direction obtained from other geological evidence is contrary to that indicated by petrofabric analysis, it is the latter that is regarded as being in error. In fact, specialists are wary about accepting general conclusions based on limited petrofabric analyses.
When writing of his studies of glacial till in Scandinavia, B. Jarnefors has used a cautionary tone:
C. D. Holmes, who studied till distribution in the central New York plain also noted that:
Holmes, who published a lengthy series of articles on this topic, including a Ph.D.
dissertation for Yale University, was especially concerned with the various conditions that "give
rise to the distribution of long axis orientation in disagreement with expected distributions".
Another investigator, R. P. Kirby, undertook studies in the Lothian region of Scotland.
This low lying area, sandwiched between the Southern Uplands in the South and the Grampian
Hills in the North, contains numerous deposits of overlying till strata laid down by glacier
advances from one or another of these two mountain ranges. Petrofabric analysis was utilized by
Kirby "as the principal tool in attempting to find the source direction of the ice sheets producing
Kirby's results are consistent with those obtained by previous workers in this field since the majority of analyses performed show that for each distinct till the long axes were strongly orientated, with the inferred glacier direction agreeing with that determined from other geological considerations. In the lowest, or basal, till, eighteen of the twenty analyses performed had a preferred orientation in a direction east to west, with the majority of dips to the west or into the flow of the ice. Two particular fabrics gave no preferred orientation but were explained by special geological factors operating at these two sites.
Fabric analysis of the till overlying the basal till showed a consistent pattern, for, of twenty-five distinct analyses, all but one showed a preferred orientation in a north-south direction. This direction was at right angles to the orientation of the basal till. The pattern of long axial dips was particularly valuable in determining the movement of the ice as in a north-south or south-north direction. As seventeen out of the twenty-four fabrics had a dominant southern dip, the indication was that the ice had probably moved from the south. Sixteen fabric analyses were performed on the topmost (Roslin) till and all but two gave a preferred orientation in a north-south direction. However, the distribution of the dips did not indicate any preferred direction.
Another major use of petrofabric analysis is in detecting the flow of streams from the
distribution of the long axes of pebbles and other deposited material. The long axes have a
preferred orientation in the direction of the currents depositing them and generally dip upstream
opposite to the direction of the river bed itself. Thus the direction and dip of the long axes should
indicate the direction of the water depositing the pebbles. But even this comparatively simple use
of petrofabric techniques has been questioned. H. J. Fraser,
Other work utilising petrofabric techniques was performed by J. S. Bibby himself and his
co-worker, J. M. Ragg, on solifluction in the Southern Uplands of Scotland.
The reader will by now have realised that conventional use of petrofabric techniques
depends on the general principle that the long axes of material deposited by flowing media will
come to rest in a direction parallel to the flow path. By interpreting the petrofabric data,
geologists can determine the direction of the glacier or stream flow - provided that there is a
strong orientation inherent in the particular fabric. Of course this interpretation must be consistent
with the direction inferred from other evidence(s). If, however, the fabric possesses negligible
orientation, then the geologist can rarely acquire any information from the petrofabric data. To
illustrate this point, consider F. J. Pettijohn's treatment of petrofabric contours,
Contours C, D, E, and F are of use to the geologist as they display a dominant orientation from which the direction of the medium depositing the fabric can be inferred. Contours A and B show random orientation, and consequently have little practical significance since the forces arranging these particular fabrics cannot be determined.
None of the contours can be used to distinguish between different types of deposits. Contours obtained from different types of deposits can be identical, and no geologist, however competent or experienced,
could determine the nature of the material forming the fabric from inspection of the contours alone. A till contour can be identical to one obtained from either a pebble deposit or a solifluction layer. The petrofabric contour can illustrate the dominant force arranging the fabric as long as a strong orientation is present. It cannot tell us anything concerning the actual material or the nature of the force forming the fabric. Other evidence, preferably from on-site investigation, is required before either the type of deposit or the particular agent responsible for it can be determined.
There is only one particular case known to this writer where a petrofabric analysis was utilised in an attempt to distinguish between two possible modes of origin for a known deposit. A detailed investigation of this specific case is of the utmost interest since it shows the limits to which petrofabric techniques can be applied.
Vast deposits of gravel occur at Corran Ferry, which forms a partial barrier to Loch
Linnhe, Invernesshire, Scotland. Several decades ago, the genesis of these gravels formed the
subject of a petrofabric investigation by S. B. McCann.
Consequently McCann performed petrofabric analysis on three sections of these deposits.
The results obtained from this analysis were quite consistent in some respects with other geological conclusions.
But, as in many other cases, there were difficulties with this interpretation.
McCann explained this difficulty by considering the steep angle of the bed of the deposits and concluded that:
McCann also undertook similar investigations at Loch Etive which, like Loch Linnhe, empties into the Firth of Lorne.
Petrofabric analysis on two pebble deposits on the northern shores of Loch Etive showed
that "the direction and angle of dip of the long axis of the pebbles ... show a marked preferred
orientation which is characteristic of stream deposits".
There were only two possible modes of origin for these gravels and one of these would implant a dominant orientation on the gravel fabric. The other would merely subject it to more random forces. It was therefore easy for McCann to discriminate between these two possible origins from examination of the petrofabric data.
These examples illustrate the considerable constraints imposed on the successful application of petrofabric analysis in situations where the method is required to distinguish between two competing hypotheses concerning the probable genesis of a particular fabric. The technique can discriminate between two alternative modes of origin provided one imparts a strong orientation to the fabric. It is then a relatively simple matter to choose the more probable mode of origin. However, Bibby's use of the technique, whereby he attempted to discriminate between a natural or artificial genesis for the stone layer at Kintraw, patently exceeds the capabilities of petrofabric analysis. It also demonstrates that he possessed little appreciation of its inherent limitations.
Bibby argued that as the contour from the man made Sheephill Fort pavement was similar
to the two contours from Kintraw, then these three contours display evidence of a compatible
origin and that this evidence "supports the hypothesis that the Kintraw platform was man-made".
The only feature that the Sheephill Fort and Kintraw fabrics have in common is the lack
of information that can be extracted from them by petrofabric evaluation. We know that the
Sheephill Fort fabric is man-made for this had been ascertained prior to the petrofabric
examination, but we have little knowledge of the forces that arranged the Kintraw material. To
paraphrase Dr. MacKie himself, "lack of orientation implies lack of evidence".
An intriguing supposition is that if the three contours in question had displayed very similar strong orientations, it may then have been possible to show that there was some probability that the fabrics had similar modes of formation (provided of course that this was consistent with other evidence). However the three contours showed little indication of any powerful dominant forces, so Bibby's use of petrofabric analysis was totally useless. An alternative consideration of Bibby's methodology vindicates this opinion.
According to F. J. Pettijohn, contours showing random orientation are common and there
must be a diverse number of geological environments capable of producing such fabrics.
So contours very similar to those from Kintraw appear regularly in nature yet little is known as to the modes of formation. The Kintraw contours are not unique, and we know little of the particular mode of genesis of this particular fabric. What is unique is Bibby's preposterous assertion that because the Kintraw contours appear visually similar to the one from Sheephill Fort, then they must also be derived from a man-made fabric. Among the diverse conditions that can produce randomly oriented contours, we find one isolated example of human organisation - that from Sheephill Fort. To argue from this that the Kintraw fabric was also man-made is arguing from the particular to the general and would likewise demand that all similar contours were also derived from man-arranged fabrics, which is most implausible.
To conclude: the petrofabric analysis at Kintraw provided no evidence of the nature of the
forces forming the stone layer so any guesses, like Bibby's was, as to the nature of these forces
are purely conjectural. By comparing the Kintraw contours to those from Sheephill Fort, Bibby
did try to provide a "simple and direct answer" to the question posed by MacKie at Kintraw.
Petrofabric analysis, however, could not provide the answer, being a very unsophisticated
technique patently unable to discriminate between a natural or human genesis for the stone layer.
Other evidence must be called for before the hypothesis that the stone layer was created by man
can be vindicated. Unfortunately for those who contend that the stone layer was the creation of
the "astronomer priesthood" of megalithic times, the available evidence
REFERENCES1. J. A. Catt and A. H. Weir, "The Study of Archaeologically Important Sediments by Petrofabric Techniques," Geoarchaeology, edited by D. A. Davidson and M. L. Shackley (London, 1976), p. 65.
2. J. S. Bibby, "Petrofabric Analysis," Philosophical Transactions of the Royal Society of London, A, 276 (1974), pp. 191-194.
3. K. Richter, "Die Bewegungsrichtung des Inlandeises, Rekonstruiert aus den Kritzen und Langsachsen der Geschiebe," Zeits. f. Geschiebeforschung VIII:1 (1932), pp. 62-66.
4. G. Lundqvist, "The Orientation of Block Material in Certain Species of Flow Earth in Glaciers and Climate," Geog. Annaler H. 1-2 (1949), pp. 335-347; R. P. Kirby, "Till Fabric Analyses from the Lothians, Central Scotland," in Ibid. 51 A (1969) pp. 48-60; W. C. Krumbein and F. J. Pettijohn, Manual of Sedimentary Petrography (N.Y., 1938), pp. 67-74.
5. B. Järnefors, "A Sedimentpetrographic Study of Glacier Till," Geol. För. Stockh. Förh 74 H-2 (1952), p. 192.
6. C. D. Holmes, "Till Fabric," Bulletin of the Geological Society of America 52 (1941), p. 1303.
7. Ibid, p. 1305.
8. R. P. Kirby, loc. cit
9. H. J. Fraser, "Experimental Study of the Porosity and Permeability of Clastic Sediments," Journal of Geology XLIII (1935), pp. 910-1010.
10. H. Waddell, "Volume, Shape, and Shape Position of Rock Fragments in Openwork Gravel," Geog Annaler XV (1936), pp. 74-92.
11. W. C. Krumbein, "Flood Gravels of San Gabriel Canyon, California," Bulletin of the Geological Society of America 51 (1940), pp. 639-706; Idem, "Flood Deposits of Arroyo Seco, Los Angeles County, California," in Ibid., 53 (1942), pp. 1355-1402.
12. T. Murray and S. Schlee in F. J. Pettijohn, Sedimentary Rocks (N.Y., 1956), p. 75.
13. J. M. Ragg and J. S. Bibby, "Frost Weathering and Solifluction Deposits in Southern Scotland," Geog Annaler 48 A, 1, pp. 12-23.
14. F. J. Pettijohn, op. cit., p. 77.
15. S. B. McCann, "Supposed 'Raised Beach' Deposits at Corran, Loch Linnhe and Loch Etive," Geological Magazine XCVIII 2 (1961), pp. 131-142.
16. Ibid., p. 134
18. Ibid, p. 135.
19. Ibid (Emphasis added).
20. Ibid, p. 138.
21. Ibid, p. 139.
22. Ibid, p. 140.
24. J. S. Bibby, op. cit., p. 194.
25. E. W. MacKie, "Archaeological Tests on Supposed Prehistoric Astronomical Sites in Scotland ," Philosophical Transactions of the Royal Society of London, A 276 (1974), pp. 170-171.
26. F. J. Pettijohn, loc. cit.
27. J. A. Catt and A. H. Weir, "The Study of Archaeologically Important Sediments by Petrographic Techniques," Geoarchaeology, op. cit., p. 85.
28. See my reply to Bibby in the Forum section elsewhere in this issue.