Burs in May 1993
Gotland is not alone with these strange marks in blocks and rocks. They are available in a few scattered locations in southern Sweden, some in Norway and a few in Finland. They are also known from France, Luxembourg, East Africa, Brazil, Fidji and as mentioned in Australia.
But nowhere is the richness of grinding grooves so large as in Gotland, where they are spread across virtually the entire island.
More than 3600 grinding grooves are known here, of which about 700 occur in the solid limestone outcrop and the rest are divided among about 800 blocks. The latter often consist of hard rock such as granite or gneiss, but also limestone and sandstone occur.
The actual grinding grooves vary in length from a few dm to more than a meter, and in width from a few cm to more than one dm. It is usually circular shape in the longitudinal direction (see Picture 2), and many are probably made with the help of some kind of pendulum device. The length of the pendulum should in most cases have been a few meters. (The Scanian grooves seem to be made by a wheel with about a meter in diameter). The cross-section of the groves' bottom usually is a semicircle (Figure 2). The depth varies from barely perceptible to more than 1 dm.
In the mid-nineteenth century P A Säve wrote about Gotland's history. He was probably the first who studied in more detail blocks with grooves, by him referred to as sharpening stones. Notes and drawings on them in his posthumous "Gotland Collections" department "Legends", are stored at Uppsala University Library. (They are not in print). Geologist Henrik Munthe quotes in his work "About Gotland's so called Swords Sharpening Stones." YMER 1933, from Säves notes following: "In the woods and fields, one finds at times large rocks, mostly of granite and also of lime stones with two, three-inch deep grooves or furrows, which go beside each other across over the stone somewhat smooth surface, these grooves can sometimes be five, six, eight nine, and are usually two or three inches wide. - People are saying that our ancestors did grind swords on them - or else that the dragon of them scrubbed his estate. - Probably these stones have been used to grind stone weapons and flint implements by our ancestors, which they call 'Thors-wedges', and there are numerous on Gotland, although not of local rock or limestone. - They lie often near the waterfront, in brooks. - Entire limestone outcrops exist, which are cut horizontally and at the side of each other with channels or grooves, " grooves learn always, although the grooves are in different places, go in the same compass."
Particularly noteworthy, according to Munthe, is the following passage after the mention of "Two large granite blocks with several deep grooves - near När's harbour": "stone axes or other weapons are believed to have been ground, because there where in the neighbourhood finds of both substances to the axes and other stone weapons, parts of such weapons, that have failed. "According to Munthe, this task is remarkable "because there is no report of finds of such waste, etc. even at places which have been described as veritable 'workshop places', such as Ronehamn."
At this locality Säve says, "Fifty-two sharpening stones available (1866) at Halor's fields, not far from the shore, or Såghagen (near Halorsåg) and on the north part "within a few acres in space, near Ronehamn." "All these fifty-two grinding stones are of granite, more o less ground sometimes to about 2 inches depth, whence (preferably when they lie in a brook), you may have to think that these furrows were made by a mans hand, by perpendiculars arm motion during the grinding process - perhaps by stone tools - seemed the major depression in their midst. "|
A W Lundberg (1873) seems to be the first (after Strelow) to draw attention to the grinding grooves on Gotland. In Vitterh .-, Hist .- o Ant.-Akad's monthly pages, he writes, among others the following: "In a pasture a few steps away from the sea-shore at Ronehamn on Gotland's southeast coast lie scattered a host of pretty large boulders of gneiss and granite, of which at least thirteen on the upturned side bear ground grooves, which in general are not parallel, but going in different directions. Some of the stones are large, others small, some have several grooves, others only one or two, some grooves are deep and long, others short and shallow, but all the grooves are apparently made by hand of men in a very remote antiquity, and not as some claim, a product of stones frozen in the ice, which under the ice in progress would have formed grooves, because in the latter case, the tracks have become almost parallel, and also along their entire length the same deepth and width, which now is not the case."-"All are deeper and wider in the middle than at the ends and have a half-round or semi circle shaped cross section and could not have been any other way than by grinding back and forth with smaller stone tools"-"on the side of the same (stone) standing workers, could comfortably take the tools in the direction the grooves meant. "
In 1918, Rutger Sernander made studies at Lake Fardume. Five grinding blocks with grooves were lying in the water, but the lake had been lowered about 1 m in the late 1800's. He therefore came to the conclusion that the grinding grooves were added during a period of dry climate, namely subboreal time, which roughly coincides with the Neolithic and Early Bronze Age.|
Figure 3. Fardume II, raä 7 Rute. Grinding groove stone15 m off shore in the lake Fardume träsk. The lake has been drained 1 m in the 19-th century.
Even geologist Henrik Munthe, who took an interest in these stones in connection with his studies of landelevation, accepted in the beginning this date.
Among others he mentioned a stone with grooves forming a stem stone in a stone ship at Digerråir in Garde. This block is now fallen, but is reported to have been established 100 years earlier. Grooves on this block would then be older than the Bronze Age. (See figure 4).
Some grinding grooves blocks are just a few meters above sea level, so from the beginning, he considered that the landelevation had not been greater since the Stone Age. Later, he did however redate the landelevation, in that he extrapolated shore levels from the Stockholm area straight into the Baltic to Gotland. Although he points out that it is highly questionable to do such an extrapolation. The redating came to dismantle the current perception that all of the grinding grooves descended from Stone Age, because the lowest levels of grinding grooves came to stay at the Iron Age water line. But since picture stones, 400 - to 900 A.D., with the grinding grooves had been discovered and he could not imagine anything but that the grinding grooves must be younger than the images on the stones, he then accepted that dating. Grooves were thus according to Munthe a total mystery, but he accepted the archaeologists opinion that it was a sort of cult.
Figure 4. Digerråir I, raä 3 Garde. The stone is one of the stem stones in the stone ship. Note the hollow to the right where the stone has stood. It has turned over to the left over a hole dug at treasure hunting in the 1800s.
When 'the learned people' no longer had any explanation for grinding grooves, the more popular notion that they encountered in grinding sword took foothold among them. In 1972 there was a fierce debate in the Gotland papers on the origon of the grooves. - Had it been used to grind a sword, and if so how? It was pointed out that sword from the Middle Ages and Viking Age (and earlier) did not fit into the grinding grooves because they are too wide. - Which of course also the grinding experiments, which were made during the debate, showed. One commentator argued that the grinding grooves were made in the 1600's because the narrow swords that could fit (and which were sufficiently flexible) did not exist until then. Of course, others pointed out that, if so, should this sword grinding be known through oral tradition.|
As stated earlier Strelow wrote about grinding grooves in the 1600's, and he thought it was a natural formation.
Carl von Linne visited Gotland in 1741 and describes in detail in his studies of Gotland's nature and culture in his book on his Gotland and Öland trip, but he mentions nothing about the grinding grooves. He ought reasonably to have encountered at least some such. - The reason for the silence on this point is probably that he could not explain their origin. (Presumably, he wanted to appear knowlegdeable in all areas and only describe the things he knew and could explain).
Figure 5. The grinding grooves block in the stone chain in a grave edge chain.
Strands I, Raä 2, Öja.
A stone with 17 grooves is in the edge of the stone chain to a prehistoric tomb in Öja (Fig. 5). The stone is located in the eastern edge with the grinding grooves on the vertical side of the stone, showing that it originaaly had grinding grooves and was then placed there at the tomb's construction. The grave is not excavated and dated, but a probable age is 2000 -. 3000 years old.
In addition, there are now so-called carbon-14 dating for a couple of blocks with grinding grooves for a lower limit of their age, indicating that these are at least a thousand years older than the time there were narrow and flexible swords. The hypothesis of sword sharpening can therefore be dismissed. - Broad swords do not fit and slender swords are too young.
Other commentators believed in other types of gear manufacturing including the grinding of stone axes. For one of the debaters, KE Gannholm, the debate was the start of a fairly comprehensive research, which resulted in several books. (First with "The mystery of the grinding grooves stones"-KE Gannholm 1974) His main thesis was that it was the fertility cult. At Gantofta in Scania are many hundreds of grinding grooves on a vertical rock wall. The location indicates that the grinding must have been very difficult, so there could not be a matter of commercial grinding but some kind of cult. Elsewhere, such as Luxembourg, the grinding grooves have awkward placements.
Figure 6. Grinding grooves at Gantofta in Scania.
Photo: KE Gannholm
My role in the exploration of the grinding grooves began in1981. Although I have a keen interest in archaeology my main focus is in the scientific direction. (I studied astronomy, mathematics and physics in Gothenburg and Lund.) The combination of the disciplines of archaeology and astronomy will be archaeoastronomy, which deals with ancient memories that may have an astronomical orientation. That is to say that our ancestors would have observed the directions of horizontal points of rise and set of heavenly bodies. Archaeoastronomy is really a subdivision of archeology, but as it requires good knowledge of astronomy it is the foremost astronomers that are engaged in it. (Since there is another category of scientists who have 'stomped' on the archaeologists' field, archaeoastronomy has been seen with skepticism by those.)|
Curt Roslund, Fil. Dr. of astronomy at Chalmers University in Gothenburg had in the 1970's studied some ancient remains from archaeoastronomical point of view, among other, the stone ship Ale stenar at Kåseberga in Scania. But already in the 1960's, I had read about Stonehenge and other antiquities in the British Isles, which they wanted to give an archaeoastronomical interpretation. - Could there be such relics on Gotland? In the winter of 1980-81 I decided to initiate archaeoastronomical research on Gotland, anyway matured. If there is enough of any kind then you can make statistical surveys. From the number of stone ships it should be enough to obtain a statistical orientation distribution. But I also came to think of grinding grooves. These are admittedly quite short and provide no long sight line. But it all depends on what you were looking for. In an apparent jumble of different directions you could maybe find a system, in which case they showed that they were not randomly oriented.
At Whitsun in 1981 my father, KE Gannholm, went to Hugraivs in Gammelgarn to see the sunset. There are grinding grooves in the bedrock and some of the grinding grooves form like a V. The sun seemed to go down almost in the direction of the tip of the V's. - It was a couple of weeks before midsummer, and he assumed that the sun would go down in the peak of the V's in midsummer.
At Whitsun in 1981 my father, KE Gannholm, went to Hugraivs in Gammelgarn to see the sunset. There are grinding grooves in the bedrock and some of the grinding grooves form like a V. The sun seemed to go down almost in the direction of the tip of the V's. - It was a couple of weeks before midsummer, and he assumed that the sun would go down in the peak of the V's in midsummer.
He went the coming days and weeks around Gotland and measured the directions of a great number of grinding grooves.|
I worked out what the extreme directions were of the sun and the moon rising and setting points, and we compared the measurements with these. It seemed as that there was preponderance of the directions E-W, N-S and around the sun's extreme northern and southern rising and setting points. Even around the moon's extreme rising and setting points the grinding grooves seemed to group themselves.
He published the first results after a few weeks, 'The Gotlandic grinding grooves and the astronomy of the old Gotlanders' KE Gannholm 1981 (DE GOTLÄNDSKA SLIPSKÅRORNA och GAMMELGUTARNAS ASTRONOMI), which of course I thought was too early. It needed a thorough survey of the distribution of grinding groves over different directions.
But as the idea of connection of grinding groves to heavenly bodies had been presented in the book the newspapers made it public. The Uppsala astronomer Göran Henriksson saw it in the paper. He vacationed on Gotland in late July 1981 with his wife. When they came to Haidby in Kräklingbo parish they stopped at 'Tors bod' (a souvenir shop) and asked the proprietor Sven Spjut if he knew where the archaeologist Johan Engström lived. He knew him from earlier and knew that he lived near the hill Torsburgen. Sven Spjut then asked Göran Henriksson if he was an archaeologist. - "No, I'm an astronomer," he said. Then Sven Spjut showed him the grinding grooves that he had behind his barn, and that someone a few weeks earlier had put forward a theory that they would have an astronomical significance. What kind of theory did Sven Spjut not know, but Göran Henriksson thought that if they had some astronomical significance it was probably something like Stonehenge.
With the help of a compass, calculator, and the Swedish Calendar, he could quickly find that the grinding grooves were directed towards the extreme north rising and setting points of the sun and the moon. At his return to Uppsala, he read the article in the newspaper Gotlands Allehanda, July 1 1981, which dealt with KE Gannholms book. He could then say that it contained exactly the same astronomical interpretation as he himself independently arrived at on July 30 th 1981. Göran Henriksson already had a computer program that could calculate when solar and lunar eclipses occurred several thousand of years forward and backward in time. He could now modify his program and with the help of it test some different principles that might be the orientation of the grinding groves. More on this below.
In 1985 I published in the newspaper Gotland Allehanda the results of my hitherto conducted measurements, as well as a simulation on the computer. The material contained directions of 801 grinding grooves as measured with the compass (488 in solid rock, and 313 in blocks considered firm and unaffected), accuracy of approximately ± 1°. A simulation of 800 randomly directions was then performed to calculate the probability if the directions of the grooves had come up by chance. |
The histogram below shows the 1253 grooves' directions in class width of 3°. Of these, 628 are in solid rock, and 625 are in blocks (assumed unaffected). All grinding grooves in solid rock have been measured with the theodolite (accuracy down to ± 0.1°) and those in blocks have been measured with a compass. The highest column, is 42 units (grooves) high, the second highest is 35 units high, then there is one column 33 units high, three that are 32 units high each, etc.
I then simulated 1000 histograms with the same number of directions and the same class width as the real one. That is, now with 1253 random directions in each class and with the width of 3 °. - The results of the simulation are shown in Table 1 below.
From this follows that in one of the charts was a column that was the height of 42 units or higher, which in turn means that the probability of any bar to be as high is 1/1000 (which is equal to 0.1%). In 165 of the charts was a column 35 units high, or higher. Thus, the probability that two such high bars occur in the same diagram is (1/1000) x (165/1000) = 0.000165 = 0.017%. |
The probability that the six highest columns occur in the same diagram were found to be 0.004%. Then it should be noted that this is the probability that there exists six that high columns in a single histogram. Note that four of these columns are the four columns between 84° and 96°. The probability of four tall columns next to each other is very much smaller. The fact that they are centered around the east-west direction means that the probability of grinding grooves would have random directions becomes extremely small.
In addition, the histogram display has an obvious symmetry, if you study it in more detail. On both sides of the central concentration around the east-west direction minima is at about 27° and 57° and with a concentration in between, and at about 114° - 126° and 150° and with a new concentration in between. - The moon's northernmost rising point vary over a 18.6-year interval between about 27° and 54° (at Gotland). The southernmost risings varies between about 126° and 153°. The southernmost settings are largely 180° from the northernmost risings and so on. There is also a concentration near the north-south direction. Grinding grooves exhibit thus in astronomical context, interesting directions.
In a chart with the class width of 3° are 60 columns. The probability that there is a peak (or a concentration at) a given direction is 1/60th. The probability of a peak at one of four given directions is 4/60 or 1/15, and a peak of each of the four directions is (1/15)x (1/15)x (1/15)x (1/15) which is 1/50625th. The reasoning is of course only approximately applicable here, but still gives a good clue.
It should be added that Gotland has an extent of more than one degree of latitude. However, it is in this context not as large that it needs to be corrected for.
If we compare the two groups of grooves in the bedrock and in blocks it will be very interesting. (Histograms in Figure 8 and 9). - From the two histograms separately, it is not so easy to draw firm conclusions. But the comparison between the two groups (along with the merged histogram), gives a very clear ruling. -
|They show that the same directions can be found in both groups (you can ignore statistical fluctuations). Now you can test the likelihood that they represent observations of the same phenomenon. The best type of correlation is linear correlation. The null hypothesis is that data are uncorrelated. The likelihood that the distribution of directions of the blocks respectively in bedrock would not be the observation of the same phenomenon was found to be 0.0646 i.e. 6.46%. In other words, the probability that they really represent observations of the same phenomenon 100% - 6.46% = 93.54%. This is a reassuring result especially considering that there is a difference in the average latitude of the grooves in blocks and in bedrock. By comparison the same thing was tested in two random distributions, both with 625 directions and with the same class width as above, i.e. 3°. There was 91.46% probability that they were random. (The probability of randomness goes to 100% when the number of observations approaches infinity). One can divide 0.0646 by 0.9146 which becomes 0.0706, and conclude that the probability of a random match between directions of grooves in blocks and in solid rock is 7% or 1/14.|
The different probabilities studies above provide the first of all note that the grinding grooves on Gotland are not randomly oriented. - As has been argued that the reason would have to do with Gotland's geological structure. Gotland has a north-south or northeast-southwest extent. It has been pointed out that many grinding grooves are located near streams and that rivers on average flow perpendicular to the coastline. This should, it has been claimed, affect grinding groves' orientation. - If it had been like that then would grooves on one part of Gotland have some directions while grinding grooves on another part of Gotland would have other directions. - On the contrary, a certain direction, e.g. E-W or N-S actually occur just anywhere on the island. Conversely, in the very same place there are grinding grooves in diverse directions such as the moon's north settings and the E-W. (The two groups of grooves in bedrock and on blocks are also in part from different areas of the island. The proportion of grinding grooves located on the rivers is also not very large.)|
If you try to submit the cause of the concentration of grinding groves around certain directions, one must consider the probability that these directions coincide with the most interesting astronomical directions is very small! Therefore, the connection of grinding groves with the heavenly bodies can be regarded as proven.
It has been argued that this connection would be that people turned in certain directions in order not to get the sun in the eyes when grinding. - But then you would definitely not get a graph that looks like this, why this claim can be ruled out immediately.
It is also argued that grinding of commercial objects of superstition did this in some directions, e.g. winds or sunrise and sunset directions at the summer solstice. - In that case, the concentration represent statistical distributions on those directions
If grinding grooves had only been one on every block or slab, you had possibly been able to draw this conclusion. But when there are many grooves on the same block or rock, these are generally not parallel, as they should be if only you had targeted them roughly against these main directions.
- If, for example you cut in the direction of sunrise at the summer solstice, it had when they started grinding the next notch been content to focus on the previous, etc.!
In reality the grinding grooves often have fan fashion, a form that is often broken by grinding grooves that cut across to the other. (Sometimes it cuts straight across another grinding groove).
In the picture in Figure 10 there are two grinding grooves that merge into a single. This should prove that they are carefully oriented in certain directions and not approximate.
Figure 10. Halor II, Raä 52, i Rone.
- It is difficult to grind on a small angle to another groove, which you hadn't done if it was not just the exact direction in which you were interested. - E.g. the moon's rising or setting point at any given moment.
- Had you only been interested in the approximate direction (e.g. if you wanted to avoid being blinded by the sun or of superstition did grind in a certain direction) you was very well able to grind the grooves in parallel. The existence of such grinding grooves is, I think, enough to prove that the main purpose of them is their exact directions. Stones with grinding grooves arranged like that are also in several other places such as at Peinarve in Levide, at Hörte and at Dahls in Lye.
I also think that grinding grooves were not for benefit, and that the grinding must have been taking place in a solid suspended device (wheel or pendulum). - To grind by hand on the edge of an existing groove would probably be difficult. In any case, one would never try to grind in the most incredibly stupid place on purpose if you just wanted to grind some object (such as a stone axe).
Moreover, the grinding grooves are often artistically rendered. They are always smooth and fine and there is no sign that angular objects had been processed in the grooves, which one should expect if they were just a byproduct of some manufacturing process.
It was the achievement of the grinding groove that was the purpose of grinding. - Otherwise you would not have made new cuts all the time despite having a good track that should have been used yet for a long time. (Grinding grooves in a group may have very different depth from about 0.5 cm to 1 dm deep mixed anyway.)
Had it been a question of commercial grinding the grooves should have been largely arranged in parallel and ground to a maximum depth (except possibly the last in each group).
I have also gone through the archaeological inventory at the National Heritage Board of Gotland and therefore I have all the material for grinding grooves collected and accurate maps. Gradually, I have personally visited the grinding grooves and measured in the first place their directions (on blocks if they can be considered unaffected). I initially used compass (standard error of the measurements about ± 1 °). Later theodolite has been used and the standard error of the measurements is about ± 0.1°. Moreover, the compass declination is avoided as the sun decides azimuths. (One can expect that those who did the grinding grooves had an average error of around 1° in the direction of the grinding groove and it is useful to have an accuracy which is in an order of a magnitude better.)|
For these measurements, I used a custom-made device that is placed in the grinding groove (Figure 11). It consists of a base plate made of iron, which is 60 cm long and 3 inches wide. On this are two 12 cm high legs on which a 150 cm long rod of alloy is attached.
On the opposite ends of the rod are two cone-shaped tips 1.5 cm high. That way you get two points as far as possible from each other. The device can usually only be in the grinding groove in one way, it locks in it).
The grooves are usually round bottom in cross section and in longitudinal section (the length is usually between 50 cm to 100 cm). The direction of each of the two cones is measured, as well as the distance with a tape measure from a solder under the theodolite. These points are then determined in a polar coordinate system. The values are fed into a computer program where the points are first converted to xy coordinates on which direction between them is calculated. The program also added a control by the distance between the tips which is also calculated. This distance is 147.0 cm and you can see, therefore, if any erroneous measurement has been made.|
In addition, the length of the groove and the instrument's placement in the same has been measured. Using these values a precise map of the area can be plotted. So far, the directions of more than 600 grinding grooves have been measured with a theodolite, almost all in solid rock (see below). Great help in my research, I have had by the computer program that I have written for the purpose.
Already in the mid-1700's, it was apparently a common knowledge that the great stone monument Stonehenge in southern England was oriented with its major axis towards sunrise on the day of the summer solstice. In the early 1900's, the English astronomer Sir Norman Lockyer (1836-1920) wrote a series of articles under the title Notes on Stonehenge, where he tried to show that the blocks at Stonehenge had been placed to mark several other directions of astronomical interest. He also studied Greek and Egyptian temples and found that many of them were oriented at the directions of celestial bodies' risings and settings. Such as the sun and moon and various stars. For example, the temple Amen-Ra at Karnak is orientated towards the sunset at the summer solstice. The temple has a major axis that is almost 500 m long. When Lockyer and his assistants observed the sun at midsummer 1891 they could only see the right edge of the sun from the temple. But Lockyer estimated that at about 3700 BCE the last rays of the full sun hit the innermost of the temple on the longest day. |
Other temples were directed towards different stellar rising and setting points, especially some stellar so called heliacal rise. - Heliacal rise is the time when a star first is seen rise before the sun. - These heliacal risings decided the time for special feast days. And at dawn when they saw the star they could launch their rituals. He found seven temples that were directed at Sirius, whose heliacal rise coincided with the annual flooding of the Nile. In general, it is not easy to know at which star a prehistoric structure is pointed. As a particular star may fit for a specific time period and another star may fit for a different time period. In the case of ancient possible connection on stars, one needs to strictly know the age of the ancient monument.
At the same time that Lockyer made his inquiries another Englishman, F C Penrose, was in Greece. As the construction period for the temples of Greece were better known than the temples of Egypt, it was easier to determine what various stars had been observed. Many temples were designed so one could observe the stars' heliacal rise - or set, i.e. when it was observed the last time in the evening after sunset.
When this heliacal rise (or set) had been observed it was known that when the sun then did rise, it would come to light up a marble statue of the god on the altar.|
After Lockyer and his contemporaries there were not so many archaeoastronomical surveys. But in 1963 the American astronomer Gerald Hawkins, brought a lot of attention when he managed to prove that Stonehenge could have had an astronomical function. And since then, studies of the possible astronomical orientation of some antiquities have enjoyed an increasing interest.
In the 1950's, the Scottish professor of "engineering" Alexander Thom (1894-1985) begun to study the monuments from the Stone Age in northwest Scotland and the Hebrides. There are hundreds or thousands of them. They can consist of single standing flat boulders, groups or rows of stones or large ring or cross-shaped plants.
Thom was a specialist in measurement technology and became interested in plant form, but also their orientation. Thom found that a lot of the stone rings were ellipses, while others were ovoid. Furthermore, his measurements indicated that the designers had used some mathematically interesting connection: the "Pythagorean" triangles, that is right angled triangles with sides that relate to whole numbers (3:4:5, 5:12:13, etc.). He thought he could even find that they used a "MEGALITHIC yard", with a length of 0.829 m, as a unit of measure.
Elsewhere in the world had been observed ancient remains from archaeoastronomical viewpoint. In Central America Mayan Indians had built a highly organized society with cities and large temples. They had a complex calendar as we know, thanks to surviving records. They made careful observations of celestial bodies and also oriented their temples in important astronomical directions.
In recent years, studies have been made in areas inhabited by some North American Indian tribes. They lived in the ages before the arrival of the Europeans. Often there are rock paintings with symbols such as a sun, which can be found at places where the sun or the moon's northernmost rise can be observed in the direction of a prominent formation in the terrain.
|north||south||max. north||min. north||min. south||max. south|
|north||south||max. north||min. north||min. south||max. south|
Stones standing in a row have a fixed orientation, as well as stone rings with oval or ovoid shape which define the orientation of major axises (the maximum diameter). A few stones can be said to be oriented in the flat side direction. |
The directions that have the greatest archaeoastronomical interest is as previously mentioned cardinal directions (E-V and N-S), as well as directions to the sun's and moon's extreme northern and southern rises and sets. See Figure 12 above, showing the eastern half of the horizon. (The rises and sets are symmetrical about the north-south direction).
The sun's path across the sky (the ecliptic) is a reflection of Earth's orbit around the sun. - It is the earth that is moving, but seen from the Earth it is the sun that moves around the sky, one round in one year. Because the Earth's axis is not perpendicular to the Earth's orbit plane but deviates about 23.45° this means that the ecliptic is inclined the same degrees of the celestial equator. (The Celestial Equator is the circle in the sky that at all the time is over the Earth's equator). - This is what is the cause of seasonal changes and thus the variations in the rise and set directions and in the height of the sun's meridian passage.
The ecliptic inclination to the celestial equator has the consequence that the sun's northernmost declination (= equivalent latitude in the sky), which occurs at the summer solstice, is just 23.45° . This leads in turn to rises (sets) of the sun in its northernmost point. How far north is due to the local latitude. It is also important if it's the sun's top edge, centre or bottom which is observed. At Hoburgen is the point where the sun's upper edge breaks the horizon line on the day of the summer solstice at about 41.2° (azimuth or bearing) and at the Northern Fårö at about 39.2°.
For each day, the sun rises farther and farther to the south, and at the winter solstice rises and sets in its southernmost point. This is approximately 134.9° at Hoburgen and about 136.5#176 at Northern Fårö. (For the settings these degrees must be subtracted from 360°).
TThe direction of the moon's rising also varies from north to south and back to north. But the hike to gain the point back and forth as the sun takes a year, takes the moon a month (moon's sidereal rotation period is 27 days 7 hours and 43 minutes).
But the moon's orbit is more complicated than the Sun's (Earth). The Moon's orbit plane is inclined 5.15° against earth orbit plane, which therefore means that it is tilted 5.15° to the ecliptic.
Furthermore the plane of the Moon's orbit is turned around one turn in 18.61 years. The Moon's orbit on the sky cuts the ecliptic at two points, nodes, and these shifts thus along the ecliptic with this period. (Nodes shift to control the solar and lunar eclipses). The moon's maximum north and minimally northern declination of the present is 28.60° and 18.30°. This in turn means that the moon usually will not have the same rising and setting directions as the sun (see Figure 13).
The above-related means that the turning points of the moon's rising and setting directions at the north the and south also moves back and forth along the horizon to the same period.
- At one instant the moon rises at its northernmost point (moon's stand still) (about 27.6° at central Gotland). |
Barely 14 days later the moon rises at its maximum southernmost point. After another 14 days it is back at its northernmost rising point, however, not quite. Each time the moon returns to its northern turningpoint, it will be shifted farther to the south. After about 9.3 years, the moon's northern turning point will be shifted to be around 54.6° at mid-island. Then comes the moon's northern turning point to re-shifting to the north (and the opposite for the moons southernmost turning point).
What has been said above applies of course also for the moon's setting points). The turning points (stand still) of the moon's risings and settings are thus (approximately) symmetrically around the east-and west point. - I.e. some years the moon's rising and setting points vary much back and forth along the horizon, while the variation back and forth 9.3 years later, is much less.
Table 4. The sun's and the moon's maximum north declinations during the late Stone Age and the Bronze Age.
Apparently, the Earth's axis tilt steadily has declined over the past millennia. That is the inherent lag of the risings and settings that is the basis of the astronomical dating of ancient monuments.
Göran Henriksson, Fil. Dr. of astronomy at Uppsala, has further developed the theory of grinding groves' astronomical connection. Figure 14 shows a map of the grinding grooves at Hugraivs in Gammelgarn, which form a continuous series of 32 grinding grooves. He has interpreted this series as a lunar calendar. Thus, he has been dating the entire series of individual years for each grinding groove.|
Figure 14. The 32 grinding grooves at Hugraivs in Gammelgarn. This is a coherent series.
Measurement of the directions with the theodolite and the rendering of the map with the help of a computer have been performed by Sören Gannholm. The datings have been performed by Göran Henriksson. The dates are day, month and year BCE.
Henriksson assumed that a particular principle has guided the grinding grooves. This can be summarized as follows: |
1. Full moon..
2. The sun at extreme points in summer or winter solstice. Or the sun is at the spring or autumnal equinox mount point.
3. A groove was ground in the direction of the moon's rising or setting each time the full moon occurs within one day from the two above dates.
4. The grooves are arranged chronologically next to each other.
At full moon the moon is brightest and the "Moon God" seems to be the most powerful and most important. People must also have discovered that it was not a whole number of moons in a solar year, and may therefore have regarded such an event that the full moon occurred at the solstices and equinoxes as a remarkable phenomenon. They had probably also noticed that the moon rises and sets variously far north and south during different years and this may have brought their curiosity and wonder. One way of studying this is to select the direction of the rise and set of the moon at any particular time of the year. The most extreme and important occasions are those stated above.
It seems, however, more natural to study the full moon's rising and setting at the winter solstice than at the summer solstice. The moon will then stay the highest in the sky and dominate the night while the sun is up during the short winter day.
- We may for a moment imagine how our ancestors in prehistoric times might have perceived the world and the universe. The sun gave warmth and light, the higher it came in the sky the warmer it became and the day was longer. The Sun God must have been more powerful than the Moon God, as the latter always tried to hide from the former. The closer the moon was to the sun in the sky, the smaller it became. In the end it was only a slim cut and disappeared entirely when it slipped past the sun. But when it walked away from the Sun the moon became larger and brighter again. Eventually it reached its most powerful form - the full moon. Since the full moon was just up all night, people came to associate this with cold. (One can still hear people who think it is cooler at full moon than at other times).
In the summer the Sun God was as most powerful, while the Moon God must have been relatively powerless. When the moon was full on summer nights, it described the very low path across the sky from SE to SW. But in winter the Sun God was powerless, while the Moon God was very powerful. And some years the Moon God was apparently even more powerful than otherwise. Then it did rise (or set) further north than usual, and was significantly higher in the sky when it was in the south at midnight. Perhaps people had the idea that the sun grew stronger the further north it came, and they did even think that the moon became stronger the further north as it came. But in the human imagination the moon did control cold weather. And then it would become colder and colder. People were perhaps even afraid that the moon would be too far north. - If it stayed there the moon perhaps did take away the power from the sun and it would be eternal snow and ice? |
- The full moon is always opposite the sun in the sky and the moon does then always rise when the sun sets and sets when the sun rises. If full moon occurs at the winter solstice it rises in NE (about where the sun rises at the summer solstice). Such grinding grooves that are targeted to NE is thus directed towards the full moon's rise at the winter solstice. And the groups that are targeted to NV is directed towards the full moon's setting point on the winter solstice.
At spring and autumn equinoxes, the sun rises and sets straight to the east and west, and the full moon also does so but is at the same time in the opposite direction of the sky (with deviations as above).- The grinding grooves at Hugraivs have an east west orientation and are therefore targeted at the rising or setting of the full moon at the spring equinox (east-west grooves could also possibly be cut at the Autumn Equinox).
The grinding grooves would thus be ground at the equinoxes and solstices. Now there's a another date for which it has been important in ancient times, namely the time midway between the solstice and equinox, i.e. the beginning of February (about Candlemas). At that time, half the winter time, and half of the winter stores are required to remain. A number of groups of grinding grooves are aimed at those points on the horizon where the full moon goes up or down in early February.
If you then select the directions to the rise of the full moon at a particular time of year, it would not result in the same direction every year. Due to the aforementioned shift back and forth along the horizon of the rise, with the period of 18.61 years, the directions will therefore have to be within a certain range (at the winter solstice between 27.7° and 54.6°, at the middle part of Gotland today).
The moon face (like full moon) occurs on the same date by a 19 years interval, (or 235 lunar months), the so-called Metoncycle. If grinding grooves have been honed and focused on the principles above, then a new groove has been ground every 19th year. During this time interval, the risings and settings shifted back and forth a full 18.61 years period and a little more, i.e. the difference between the 19 full years and 18.61 years.
- If at the winter solstice, the full moon occurred a year when the moon was at its absolute most northerly point, the full moon will then 19 years later rise in a little more southerly direction. 19 years later in the slightly more southerly direction and so on. Grinding grooves directed towards these directions will then be part of a fan shaped pattern. |
The difference between 19 years and 235 lunar months amounting to almost 1.5 hours and makes the Metoncycle derail after a few centuries. A new Metoncycle will then start and for a while you will have two Metoncycles alternating. The joint between two Metoncycles will then overlap each other for some time. Therefore, it is sometimes eight or 11 years between the times that the full moon occurred on such winter solstices, and that they therefore have ground grooves at those occasions. The moon's rising point is controlled by the 18.61-year-period, which will result in the fan shape being broken and a grinding groove cut obliquely against the other. This is often the case.
Around the year 3000 BCE the full moon passed the star Antares in Scorpio on the day of the vernal equinox. This star could be used to mark the vernal equinox. - The moon passes Antares course every month, but it's only once a year as it can be a full moon at this passage. But usually it is therefore as much as 19 years between the times when the full moon happens the same day of this passage.
During the same period, i.e. about 3000 years BCE, the star Spica was in Virgo on the part of the sky that was passed by the full moon in early February. Similarly, one could use this star to mark the 'mid winter.
Because of the so-called Luni-solar precession, which depends on the solar and lunar impact and the consequence that the Earth's axis rotates slowly around in space - a lap of 26,000 years -. This in turn leads the stars to adopt new modes in the sky. The full moon did pass Antares at progressively later dates. Antares has thus gradually moved away from the vernal equinox. But people did apparently stick with this star as a marker for several hundreds of years, until this moment occurred one week after the vernal equinox.
Henriksson's method to date the grinding grooves is to figure out the directions of the moon and sunset at these very special occasions, seen from that point. The computer calculates backwards in time and a series of grinding grooves are compared with those values. If, for any period there is a coincidence (in a few degrees) between the average grooves direction with the calculated values, then the entire group is dated. It is certainly the case of periodic processes and a repetition of a pattern should be possible. But apart from the Luni-solar precession there will also be a change of the slow tilt of the Earth's axis over millennia due to the aforementioned planetary precession. The tilt has ever since the Stone Age steadily declined. Therefore, the moon's rising points shifted further north and south during the Stone Age than at present, and an unambiguous dating of grinding grooves is possible.
My measurements of the directions of the grinding grooves do support Henrriksson's theory. The concentration around the east-west direction corresponds to the full moon at the vernal equinox (Antares). The two smaller concentrations on each side of it do correspond to the full moon in the middle of winter (Spica). And the two concentrations further away do correspond to the full moon at the winter solstice.
Approximately a dozen grinding grooves stones are known, that are at the same time picture stones or stones with the shape of a picture stone. Archaeologists have previously taken for granted that the grinding grooves would have come after the stone was carved into the picture stone. But as more and more indications have emerged that the grinding grooves are much older, than the picture stone time, this may not be true.|
Many people who see a picture stone with grinding grooves are apparently irritated and see it only as a degraded picture stone. - I am researching the grinding grooves and thus have seen many such, I will then put in place something different from the pattern. What is most striking is that the grinding grooves on picture stones seem to be broken (see Figure 15), therefore, they former must have been longer but have been cut off.
Figure 16. One of the ends of a groove in a grinding grooves block in longitudinal cross section. From Häffinds IV in Burs. Grinded grooves are concave, but in the end they become convex.
Detail from figure 2.
Many grooves on picture stones are located with one end at the edge of the stone, so that only half of the grinding groove remains. On an ordinary grinding grooves stone, there are never a groove that has its deepest part at the edge of the stone if not the stone is broken off. Grooves are broadly in the middle of a stone so that the deepest part is there.
After examining a number of grinding grooves with respect to shape, I have observed: that almost all grinding grooves are convex at the ends but concave in between. This is true also when the stone is shorter than the groove and one or both ends of the groove is at the edge of the stone. See Figure 16. This must be because when the grinding tool comes to the end of the groove it angles a bit and grind more at the ends. - Thus, one can directly determine if a groove is interrupted or not. In a few other loose limestone blocks there are the remains of just the tip of some of the grinding grooves on them. - And then it is not required much thought at all to realize that the grinding grooves have been broken off.
At Gällungs in Väskinde is a limestone rock aggregate as a bridge over a brook. It contains about a dozen grinding grooves. Many of the grinding grooves have one end at the edge of the stone and a few leave only the tip, which quite clearly shows that this rock has been broken up after the receipt of the grinding grooves, and that these grooves have been cut off. Even in the cemetery at Mästerby church there is a stone with such grinding grooves.
Figure 17. The stone at Sjonhem's church.
Kyrkan I, raä 15.
On the north side of Sjonhem's church is a finely hewn rectangular limestone block with a circular hollow in the middle. Across the stone is a large fine grinding groove. The casual visitor may get the impression that the grinding groove was ground after the stone was cut. But closer inspection reveals the tip of another grinding groove at one edge of the block (Figure 17). And then the grinding grooves have been there before the block was cut.
If the grooves on the picture stones had been ground after the stone had been made into a picture stone, the grooves at the edge of the stone would have turned convex at their ends. But no trace of such is observable. (I have studied a number of such stones on both Gotland and in the State Historical Museum in Stockholm).
There are picture stones, that also have stab marks in the groove where it encounter the edge of the stone. One example is the stone on Sanda cemetery, Kyrkan I, Raä 34. Stab marks have been added to an existing grinding groove and apparently when the stone was carved into the picture stone form.
18. Picture stone at Sanda church. Detail. Raä 34. Detail
At Tomase in Lokrume parish is a limestone rock with grinding grooves, which is broken up but not towed away (Figure 19). Beside this rock are four grinding grooves in solid rock. Loose limestone outcrops with grinding grooves are found here and there, among others at churches.
The stone at Tomase shows that at some time in history rocks had been broken up that already had grinding grooves on it. The Tomase-stone might well have been intended to become a picture stone.
Figure 19. Tomase IIII, Raä 45., Lokrume.
The picture shows the uplifted limestone rock with 12 grooves.
In the bedrock at right are still more grooves.
It has been argued that the picture stone masters could not accept the raw materials for their stones, "which was already destroyed by the grinding grooves in arbitrary directions." On this, one can replicate, "Never in prehistoric times have humans created a picture only because it was decorative, or had a value of beauty in the way we perceive art." (Quotes from Göran Burenhult's book "Reflections of the past", 1986). - It has always been magic involved. And yet at the picture stone period, 400 - 1000 CE, a man's imagination was, of course, largely dominated by magic and cult. Grinding grooves must also have been to them something unintelligible, as they were thousands of years older. Perhaps they believed that the grinding grooves were traces of the gods, and then you could very well have broken up the limestone outcrops with grinding grooves to cut them into the form of a picture stone because they believed that they had great magical power.
Figure 20. Stone from Hogrän, Byssegarde I , Raä 50, which previously provided a stove. The grooves' placement shows that these were from the beginning. Can the crescent-shaped depression to be a rock carving?
Of the approximately one dozen of the known picture stones with grinding grooves, only two have some pictures - and are then really just picture stones. The large stone from Sanda, now in Gotland Museum's picture stone hall (dated to 400 A.D.), has some grinding grooves at the bottom, and a few shallow grooves or just grindings on the surfaces in direct contact with the images. On the right side the border seems to be polished. Those who stand at some distance could possibly get the idea that the border has vanished at these places. But if one looks more closely, one finds that there has never been a border there. It simply stops and then continues on the other side of these polished surfaces (see Figure 21). |
FFigure 21. The stone from Sanda.
Now in the Gotland Museum's picture stone hall.
The images on picture stones were painted on the surface of the stone. The paint is since long time gone and what you see is the contour. The contours where stabbed into the stone, but only where the stone was uneven, i.e. rough. That's why you do not see any images on the smooth surfaces. The same can also be observed around the horse and around the ship. The contours have not been polished away. In some places the image contours have been stabbed into the polished surface such as the rudder. One can thus see quite clearly that the grinding were on the stone before it was reused as a picture stone!
On the other famous picture stone with both grinding grooves and images is the stone from Othemars in Othem (dated to 800 A.D.), there are only pictures on some of the stone, and this has been interpreted as the images on the rest of the stone have been polished away: Arne Philip with the article "Missing picture stone rebuke", in Gotländskt Arkiv 1986.
I had not seen this picture stone when the article was published. However, during my research on the grinding grooves I became more and more convinced that these are considerably older than the Iron Age. I had to immediately investigate it. Therefore, I went in the middle of the worst snowstorm in January 1987 to the Gotland Museum's store close to St. Hans church ruin and studied the picture stone in a wooden shed there. When I examined it I found, however, that the pictures seem to be stabbed into the polished surface. Sometimes they interrupt a contour line of a raised ground area, and continues on the other side.
Sometimes there are remains of large or small polished surfaces as islands in an otherwise stabbed area. In the case of a grid in the sail, some squares are grinded, while others bear no traces of grinding. |
- Where there are no visible traces of grinding, the surface appears as perforated, as if someone with a small tool wanted to knock out the grinding. It is impossible to grind on an uplifted surface (as the squares in the sail) and at the same time escape the grind of the next. - The inner surfaces are almost at the same level. The tracks from the grinding go straight ahead toward the stabbed contour line and then stop there, and do not continue on the next surface. This must mean that the stone was never completed with pictures. The stone appears not at all weathered and the edges are sharp. - Only a small portion of the stone shows traces of weathering, and wit both image contours and grinding grooves. The intersection between the grinding surfaces respectively the grinding grooves and the edge of the stone indicates that it is cut into a surface that has already been ground. Nowhere can be seen that the images have been cut away. This means that the grindings on the stone from Othemars are older than the images on it.
I published the results of the study in Gotland Allehanda on January 28, 1987. In the summer of that year Göran Henriksson and I investigated the stone again and Henriksson wrote an article, which he published in Gotländskt Arkiv 1988, titled "New studies of picture stone from Othemars in Othem".
The history of the stone from Othemars as I understand it: In the Stone Age there was grinding grooves in a solid limestone rock. At the same time the whole area around the grooves was polished. After not too many decades, a large piece of the appliance was broken off and was then turned upside down. Then they continued to grind on the previous bottom. - There are some blocks with grinding grooves on multiple sides, which must mean that they wanted to use a stone of very great magical power. Three or four thousand years later, in about 800 CE, they chose this stone to be cut to form a picture stone. Work on this was stopped early though.
Something else I noticed: In some of the grinding grooves on the picture stones are narrow depressions or inscriptions that follow part of the groove's bottom. I have not seen these carvings in other grinding grooves. If these carvings in the bottom of the grinding grooves appear only on the picture stones, there is no other possibility than that it is the mason (who made the stone into a picture stone) who have achieved them. It is very unlikely that a phenomenon of this kind would occur by chance in only one category of grinding grooves stones.
A few isolated cases of odd grinding grooves are also present. On Eke Church are some short and narrow grooves that cannot be with the usual making. A baptismal font from Halland has two grinding grooves-like formations. They need not be true, but it may be that the picture stones people had chosen a stone with existing grooves. A kind of stone that maybe has been used for grinding corn from Endre has a possible groove on it, Hanes I Raä 30. The original function of this kind of stone was probably to sharpen stone axes. Furthermore, single look-alike grinding grooves probably occur from very different times, made with different techniques and not by the usual way of grinding grooves.
TThere are possibly two findings of used objects in a grinding groove, one from Eksta which I have not had occasion to examine more closely. The second find is from Lau. See figure 22. The grinding grooves stone with the stone axe look-alike object was found back in 1952 at the cultivation of arable land in Lausmyr, a former swamp. |
In the debate on the grinding grooves in the 70's, it was, however never mentioned. It wasn't publicly known until I wrote about the discovery in GA in June 1988. The block, that was quite large, had only one edge of the surface visible. When the stone was uncovered, grinding grooves were discovered, and in one of the grinding grooves was something that resembled a stone axe. Unfortunately the block was blown up, but a piece of it was taken care of, as was the 'stone axe'. The block had originally about five grinding grooves, but the piece that remains is just the grinding groove in which the 'stone axe' was found. The object is apparently broken, but it was found that way.
Figure 22. Artifact found in situ in a grinding groove.
Gannor V, Lau.
The fit seems perfect between the object and the grinding groove, so there can be no doubt that it really has been ground in the grinding groove. - But are grooves the result from making stone axes? - I do not think so. As KE Gannholm pointed out, we never see scratches after angular stone material in the grinding grooves. The object is no stone axe, but just an elongated "rock sausages". If the grooves would just be the result of a manufacturing process why would you then as I said, abandon even shallow grooves, although these could have been used for a long time? Whatever you ground, you must have it polished smooth. However, it is of course possible that you may have finished grinding in a grinding groove. -That you believed that such stone axes had been given magical powers, if they were given a final treatment in a grinding groove. Grinding groves' dating to the Stone Age are supported by this finding. Although a different circumstance surrounding this finding demonstrates the Stone Age: The grinding grooves were approximately 30 cm below ground level. But even in the early 1900's the water in the swamp was in Lausmyr at least half a meter high at the site, even in the middle of summer. Here we have a parallel to Fardume swamp, see above, where Rutger Sernander 1918 did his surveys (dating: subboreal time).
Figure 23. The photo was taken when the block was unveiled. In one of the grinding grooves were the object in figure 22 found. The grinding grooves stone was about 30 cm below the surface of the mire. Photo: Stig Dotes.
According to the investigations of grinding grooves on picture stones that I explained in the preceding pages grinding grooves can be dated back to before the picture stone era. Also some C-14 datings give some results in this direction. At Dibjärs in Hörsne the brook recovered two grinding grooves blocks inside the brook's bank at an unusually severe flood in 1985. Three C-14 samples were taken, partly from the material closest to the grinding grooves, and a little higher up. The peat closest to the grooves gave the result (converted to calendar years) 585 ± 90 CE. Wood pieces and lumps of charcoal a little higher up gave (converted to calendar years) 1075 ± 75 CE resp. 1055 ± 140 CE. At Smiss in Hemse was found a grinding grooves stone with grinding grooves down at the bottom of a wall of stone. The stone was directly on a carbon pool from which the two C-14-samples were taken. They gave the results (converted to calendar years) 655 ± 100 CE resp. 635 ± 105 CE. The grooves were of course long been out of service when the stone was put in the wall.|
Start date for the grinding grooves tradition Henriksson puts at about 3300 BCE. No astronomical dating's have been demonstrated for the period after 2000 BCE. The question is whether the end date for grinding grooves can be set to 2000 BCE, or if grooves on the low level of the sea show that it continued into the Bronze Age. Added to this is the issue of shoreline displacements that I will comment on later.
Grinding grooves would thus belong to Neolithic and largely the Pitted Ware culture. The number of known grinding grooves is just over 3600 and the total number may have been twice as large, but this is impossible to know anything about. Inger Österholm "Settlement pattern of Gotland in the Stone Age", 1989, enter the total number of individuals on Gotland at this time to 500 as a minimum, and seems to think that the real number was not much larger than that. Lars Bägerfeldt "Neolithic on Gotland", 1992, is criticizing Österholm's conclusions.
Bägerfeldt writes that, so far, the archaeological research defines reasonably well the objects and tools that were manufactured as a minimum, how they are practically used as a minimum, which in the community who used them is a minimum, the amount of knowledge that existed in society at least, how many people it was per unit area as a minimum, the amount of possessions they had as a minimum, the standard of housing they had at least, how the hierarchical structure looked like at least, how extensive and complex world view and cosmology was at least, etc. The conclusion is that the real numbers was not just larger, but often probably much greater.
Österholm has at her estimates of the number of individuals assumed, among other things the number of known Pitted ware premises and the number of known graves.
Bägerfeldt shows that the dark figure in these calculations is very large. The population's real number should instead be soaring over that Österholm calculated. Bägerfeldt has made his own calculation, where he among others has been based on pollen diagram change and the relative size of certain plant species. A pollen diagram from Lojsta parish shows that all species that indicate animal husbandry and farming is at a fairly constant level from 2400 BCE to 1000 CE. i.e. the same amount of food was produced and thus its population throughout this period has remained fairly constant. Population size 1000 CE based on the number of farms is estimated at about 15,000 inhabitants. (Carlsson, A.1983).|
At grinding grooves starting 3300 BCE indicates Bägerfeldt that there were at least 6000 inhabitants on the island. At its final, population should be approached to 15000. Thus we get a better match between the number of grinding grooves and the number of people who may have been responsible for their creation. These people must also have lived in a much larger area than shown in the number of known Pitted ware and other venues.
I have not investigated what relationship exists between grinding grooves premises and other antiquities. It can be an archaeologists' task. But through my studies, I have some ideas. For example, it has sometimes been suggested that there would be some relationship between the grinding grooves and Iron Age settlements. But already Lithberg showed that the Iron Age settlements are often a continuation of settlements back from the Stone Age. However, I have the impression that grinding grooves premises are not spatially related to some other ancient remains. They are located at the old shore or by the former lakes or hills, or simply in places where they had a clear view!
In Tanumshede and Tossene in Bohuslän there are a dozen respectively twenty grinding grooves on rock premises. Accordingly to task, there is the so-called transitions, which show that the grinding grooves are older than the rock carvings. As mentioned earlier, there is grinding grooves in other parts of Stone Age contexts. In addition, there are sometimes such situations that there can be no question of commercial grinding, but cult. The grinding grooves premises are simply places of worship!
There are also some grinding grooves at the bottom of brooks. Brooks move, however, back and forth over the centuries and millennia. And earlier swamps get filled up and can be crossed by rivers. The grinding grooves stone Kvännänge I Raä 101, in Burs parish, which is now on the bottom of a brook, has probably not been ground in this position. It's probably just a coincidence that the river currently has unearthed this. In the ground around the river, there may be several blocks of the grinding grooves. At Dibjärs in Hörsne the brook dug out two grinding grooves blocks at an unusually severe flood in1985.
Ancient memories about altitude above sea level has long been of central importance in terms of dating shorelines. Ancient memories of altitude has been counted as a percentage of LG (Lithorina-border) and also directly used as dating. The work that has been the basis for this approach is what Henrik Munthe in 1933 published in the journal 'Ymer', 'About Gotland's so called Swords Sharpening Stones'.|
Prior to that time, as I mentioned earlier, had it been assumed that the uplift of Gotland since Neolithic times had not been more than a few meters. But in 'Ymer' 1933 Munthe makes a comparison between the uplift in the Stockholm area, which was comparatively well-known and time determined by Granlund, and on Gotland. Munthe writes: 'I have therefore taken from his [Granlund's] land elevation curve for the Stockholm area ranging from the Bronze Age beginning to the present and at the fig.  and introduced partly this and partly in respect thereof corresponding curves for the north central and southern Gotland and the latter entered characters for these lowest lying grinding blocks. The result is, apparently, to those in use at one stage, falling at the about 600 CE i.e. approximately the middle of the Great Migration. Nevertheless, that the aforementioned grounds footed timing to be regarded as quite uncertain, since the uplift in the Stockholm area on the one hand, and on Gotland on the other, yet can be proved to have proceeded as parallel and regularly, as the curves setting forth. 'The lowest grinding groves' elevation was measured carefully by Henrik Munthe among others at the beginning of the 1900's, and he found that they were about 2 m above sea level on the southern part of Gotland and slightly higher in the north. Munthe's article in the journal 'Ymer' thus led to a total re-dating of grinding grooves from the Stone Age to the Iron Age. Munthe continues: 'The alleged dating of the lowest lying grinding grooves blocks on Gotland has won confirmation by the findings of the grinding grooves on the picture stones of some among the island's distinctive picture blocks, which as we shall see, originate from the latter part of the Migration Period.' He adds: 'Grinding grooves on them are obviously of later date, how much later, however, is at present impossible to say.' - He was like everyone else at the time, unable to realize that grinding grooves could be older than the picture blocks.
It was our ancestors' prank during the picture stone time to break up rocks with grinding grooves and chop them into picture stone form, which confused such researchers as Henrik Munthe. This also led to accept it by him made extrapolation of the uplift from the Stockholm area out to Gotland.
Excavations of the Stone Age settlements, especially Pitted Ware, in recent years have shown that these have been washed over by the sea several times. Previously the archaeologists stopped digging when they came down to a sterile layer of sand (sea floor). But at an instant someone dug through it and met new cultural layers below. The settlement had thus been exposed to a transgression of the sea. This suggests that the shoreline displacement process was not as simple as previously assumed.
Rhodes W. Fairbridge has over a number of decades been studying sea level variations during the time period after the last ice age in a large number of locations around the world. He puts these variations related to climatic changes in global climate. In Figure 24, I have drawn from his charts of ocean variability.|
10,000 years ago, i.e. 8000 BCE, the sea level was about 35 m lower than at present because the water was still tied up in land ice. Since then the sea level was raised in stages when the land ice melted, and fell temporarily during times of colder climate. Shortly after 5000 BCE the sea level reaches the current level and remained for quite a long time up to 3-4 m above the present-day level, which coincide in time with the postglacial thermal time. For a few short periods of time the sea level, however, sank by perhaps 5 m, so that it was lying a few meters below the contemporary.
The whole Scandinavian Peninsula including the Baltic Sea, the Baltic countries and Finland has been icy, and depressed because of the weight of the ice. When the ice melted, the land raised again quickly at the beginning, and then at successively lower speed. Shoreline displacement has thus been an interaction between sea surface variations and the soil upward. The Baltic region has further complicated the picture because the Baltic Sea at certain times has been constricted off from the world's oceans. We see traces of past major subsidence in the form of the Ancylus-border wall and the Lithorina-border wall.
After the time when the Lithorina wall was formed, about 4000 BCE, the world's oceans have according to Fairbridge continued to rise and sink, but eventually with lower and lower amplitude. But how did the ground move? The most common assumption is that it has risen steadily, but with exponentially decreasing intensity. - But can we really make that assumption?
Gotland is located in the Baltic Sea and then we should consider that the displacements at the shores of Gotland possibly have no connection with the situation at the surrounding mainland. Gotland is a part of the seabed and must therefore move in the same way as this. When sea levels lead to greater weight on the bottom, then the seabed is falling and vice versa when the sea level is lowered. Ground surface movements do not in any way have occurred at a steady pace, nor need it have happened the same way all over the island. This relationship is also implied by recent research. Ground surface movements on Gotland may thus have been faster at certain times and slower during the other times or even have been reversed.
Munthe tried to use the grinding grooves in order to determine displacements of shores. Maybe we can do it now in the case we can date the grinding grooves in another way.
|From my survey of Gotland grinding grooves, I have tried to group the grinding grooves by parishes for height above sea level. I have been using the topographical map with a scale of 1:50 000 where level curves for every 5 meters is included. Then I counted the number of grinding groves stones from 0-5 m and from 5-10 m and so on. The difficulty has been to establish the blocks' original location after they have been moved. I have different categories of grinding grooves blocks. Blocks in the original position and attached to the ground. Blocks in walls at the fields I count to a category that has been knocked out of alignment, but that has not been moved far. But we have the category of grinding grooves in gardens and museums. In many cases, there is an indication of the bargains site. In most cases, there are also blocks moved out from the fields and back to the farm, so they are likely descended from the same altitude as the farm. In the case of Gotland inland, this is higher than 20 m above sea level, and both blocks original position and the current location at the farm harbors therefore above this level.||
Table 5 summarizes the material. I have divided it into two groups. I am covering the 18 southern parishes and the rest. For each height interval, I have also summarized the south and north part independently. My division of Gotland in a southern and a northern part is based on some consideration and includes parishes south of a line from the När-Lau border and to the Sproge-Eksta border. In the case of low-lying blocks perceives a clear distinction in this line.|
The lowest-lying grinding grooves stones are located at approximately 1.5 m on southern Gotland and 3 m in the north according to Munthe, but also according to by me and Göran Henriksson made researches. On the southernmost part of Gotland, it is possible that there are grooves at even lower levels. Outside Rivet in Sundre is a report on a block on the seabed at 0.5 m depth and drained at low water.
According to Table 5 there are 766 grinding grooves in the southern part below 5 m, while only 136 on the northern.
|0-5 m||5-10 m||10-15 m||15-20 m||>20 m||Undef.||Sum|
Göran Henriksson dates grinding grooves to the Stone Age. - He places them about 3300 BCE to about 2000 BCE. If this dating shall apply to all grinding grooves, even those who have a low position in relation to the sea, then the dating will conflict with many archaeologists' perception of shoreline sequence. Although the possibility exists that there is grinding grooves from very different times. - And Henriksson's datings mainly concern the grinding grooves in solid rock, which are high relative to the sea. But he has also tried to date the low-lying blocks. Because it is less grinding grooves per block the dating is possibly more uncertain. But even here he has datings in the same period. Many people have difficulty accepting a beach level of 2 m above the present before about 1000 BCE. After that time it seems that the beach has remained fairly constant with a little commuting up and down. - In the vicinity of the low-lying blocks at Häffinds in Burs parish and at the same level, 2-3 m above sea level, is a typical Bronze Age burial cairn (fig. 25) that for some reason not previously has been noted. The sea must have been lower than this because it is located on a dune with a lower area inside. It cannot have been built on a sandbank in the sea. - The tomb is not studied. But the most important factor to consider ought to be to determine the ages of low-lying graves!|
The dating of shore levels that archaeologists are using is probably the highest rates in each period because they are usually the floodings of such settlements. The real sea level was at least a few meters lower, as these floodings represented the extreme high-water installations and onshore winds.
How high the water under such conditions can reach depends on coastal topography and wind. Presumably, the variation in the Baltic Sea (between high water and low water) was different in different climatic stages. In addition, archaeologists' datings presumably do not catch more short-term changes of sea levels. According to geologists, it was high water level changes up and down in the Baltic region during the period 3000-2000 BCE.|
Grinding grooves represent on the other hand, the lowest water levels during the period. Because they were most likely in the water when they were grinded.
In summary, over the past approximately eight thousand years, the global sea level has remained around the current, but (according to Fairbridge) oscillated up and down several meters in each direction. People have also become aware that many increases and decreases of water level at Gotland have occurred. - Maybe both increases and decreases in the soil itself also occurred. The process has certainly been very complicated at Gotland. According the Gutasaga (as written down in the 1200's): "Gotland was first discovered by a man named _ieluar. At that time Gutland was so bewitched that it sank by day and rose up at night. That man, however, was the first that brought fire to the island, and afterwards it never sank again." Perhaps behind this legend is a true tradition of Gotland that in different periods sometimes rose and sometimes sank.
On May 23 1991 the archaeologist Anders Carlsson and Göran Henriksson, Fil. Dr. of astronomy at Uppsala, met in a debate on the age of grinding grooves' at Uppsala University. Tore Gannholm was present and wrote an article published in "Gotländskt Arkiv" 1991. The next section has this article reproduced.
In view of the fact that the archaeologist Jonathan Lindström at the University of Stockholm had written an essay in which he criticized Göran Henriksson's, Fil.Dr. of astronomy at Uppsala, for his dating of the Gotlandic grinding grooves, the professor of astronomy at Uppsala University, Bengt Gustavsson, had invited the archaeologists from Stockholm to discuss the topic on May 23 1991. |
In his paper Lindström suggested that many of the grinding grooves uplift would have been up to five meters under water if Göran Henriksson's dating to the Stone Age was accepted. They relied, like many others before, on Henrik Munthe curves of the uplift. Lindström argued that the Professor of Quaternary Geology Lars Königsson had told him that the grinding grooves would have been five meters under water in the Stone Age.
Against this background Fil.Dr. in Archaeology Anders Carlsson presented a theory that the grinding grooves can not be older than the Iron Age and polished with some type of wheels to get the perfect rounding. It was found that the grinding grooves are not, as previously suggested, been used for sharpening swords. Carlsson emphasized K.E. Gannholms important efforts to kill the myth of sword grinding and dismissed this as both theoretically and practically impossible.
Göran Henriksson revealed a number of examples that indicate that the previously established uplift theories of Gotland cannot apply. (Among other things, was a Bronze Age cairn recently pointed out by Sören Gannholm (GA April 16 1991), and Iron Age tombs in the Bandlunda Bay, which lies close to the current water line.)
The uncertainty of the uplift theory was confirmed with great power by archaeology professor Bo Gräslund, who was present.
Gräslund and Königsson had discussed this for many years and their views are not consistent with Lindström. Lindström has to be the victim of wishful thinking.|
Gräslund also confirmed that he the same day had been in contact with Königsson. He had then stated that the Gotlandic uplift does not follow current theories and that it is little explored. Gotland has no basement, but is part of the Baltic Sea and therefore cannot follow the same rules as for mainland rock. However, Königsson could confirm that the areas where grinding grooves are found may well have been above water during the Stone Age when grinding grooves are deemed to have been ground. The water situation in the Baltic Sea changed much during the relevant time, which makes the picture so incredibly complicated. Transgressions have a major role in it. The Pitted Ware settlements, located on the island of Gotland, which is also used as evidence of the shoreline position, clustered according Gräslund around the transgression tops, i.e. the highest level the sea can reach at "sea level rise" or flooding.
It was also suggested that you would reverse the thinking and use the grinding grooves as the basis for determining the water level of the sea during the Stone Age.It was also suggested to reverse the thinking and use the grinding grooves as the basis for determining the water level of the sea during the Stone Age.
With the above observation about the uplift falls the argument that the grinding grooves must be from the Viking Age and we are back to Göran Henriksson's dating to the Stone Age and can continue a constructive scientific debate.
IIt must be seen as very positive that Anders Carlsson has initiated a dialogue between archaeologists and astronomers in a true interdisciplinary spirit and I hope this develops into further fruitful work.
Some attempts to mimic the old grinding grooves have been made. Some years ago Axel Larsson at Hallbåtar in Levide parish made a motorized swing (to avoid having to stand and pull it by hand). On this he fastened a stone and let the machine stand and grind with a certain pressure on a limestone slab. The result looked like a grinding groove. The device was a simplification, but it took note of the fact that most of the grinding grooves had the 'pendulum shape'. Most grinding grooves are also circular in cross section and to obtain this form they had to turn the grinding object in different positions. Grinding grooves with 'U-shape' exists however.|
In the mid 1980's grinding experiments were performed by Helge Hedlund at the car museum in Vibble. He used a device with a handle. The device looked like a rocking chair's rail. On this, he attached an iron band. The idea was to sharpen swords. Why he in this way would grind away a lot of metal is unclear. (Swords are forged and banked up with multi-layers of twisted metal rods according to the book "The Vikings in the East and West" by Magnus Magnusson.) Hedlund had in any case, a result that looked like a grinding groove. But because he knew what a real grinding groove looked like, he geared himself to get the groove so smooth and fine as possible. According to his hypothesis grinding groove was, however, only a by-product and he should not have bothered about how the grinding groove looked like. I pointed out at the first opportunity some bumps and at the next occasion, the bumps were polished off! Before trying to figure out which technology is used one should have an idea of how old the grinding grooves are. When Hedlund made his grinding trial, it should then have been clear to all that grinding grooves existed long before swords, why the attempt were futile from the start.
In June 1992 the Burs Heritage Society manufactured a machine designed to make the attempt to grind grinding grooves. The construction was set at the Heritage's museum, and performed by Arne Nilsson and Mårten Wessberg, after a drawing by Göran Henriksson. The author participated to some extent and some people were invited.
The device consists of a wooden frame on which a pendulum is fixed. The pendulum consists of two long wooden staves. The upper end is attached to a horizontal log, which can roll back and forth in its suspension points. The bottom of the pendulum is fixed to two horizontal logs. Thus control of the pendulum was obtained. The pendulum has a counterweight because the two staves are nearly twice as long as was needed, and at the top are two horizontal logs. In the two bottom logs are holes drilled through which a third stick can be turned down. At the bottom of this rod is a "shoe" loosely attached. At the bottom of this "shoe" is grinding object loosely attached. When pounding down the wooden rod, it will press the "shoe" to tie the object to the surface. In grinding experiments during mid-summer in 1992 a 2 dm long piece of wood was used as grinding objects. But you can also use grinding objects of stone like the find from Lausmyr swamp, which has been described earlier. Finding of an artifact in a grinding groove
The actual grinding was performed on a limestone. After about four hours a 1 cm deep and 47 cm long, grinding groove had formed. Water and fine sand from the beach was used for the grinding. The grinding was pretty easy, once you had got into a steady rhythm.
Now and then you got the wooden rod that holds the grinding object on the ground to be knocked down just right. - If you knock down to hard it squeezes. Grinding object may periodically be removed and rotated so that different sides will be ground. In this way, grinding grooves will be given a circular shape in cross section.
Fig. 26 Slipskårependel bygd 1992. Sojdsmuseet Burs.
At the ends of the groove the grinding object will angle a little so that a convex form was achieved.
If you have achieved a grinding groove it is not thereby proven that it was that way it happened. But you have a hands-on experience and can reflect further on how it probably could have happened. There are variations in the real grinding grooves' design, which shows that a different technique can have been used for these. Grinding grooves of similar form may be found in more contexts and they have nothing to do with real grinding grooves.
Grinding grooves (polissoir) in France.
Fig. 27 and 28 . Dolmen de la Pierre Plate, France. Grinding grooves on the bottom of the roof blocks in neolithic megalithic grave.
Polissoir de la Pierre Plate
Ala Ollaivs I|
Anga Liffride I
Ardre Brändu I
Buttle Altaime I
Buttle Häxarve I
Endre Fjälls I
Fole Bondarve I
Fårö Lansa III
Fårö Ödehoburga I
Fårö Ödehoburga V
G-garn Hugraifs I
G-garn Hugraifs II|
G-garn Mattsarve I
Garde Kyrkan I
Garde Nygårds I
Garde Nygårds IV
Gard Robbenarve I
Hangvar Austers II
Hejde Haidgarde I
Hejde Lambräum I
Hejde Sigsarve I
Hejde Simunde I
Hejde Stenstu I
Hellvi L Ihre I
Hellv , Lärbr St Ihre I
Hellvi St Ihre II|
Hellvi St Ihre III
Hellvi Surrgarde I
Hellvi Surrgarde II
Kräklingb Haidby I
Kräklingb Kyrkan I
Kräklin Stabbäng I
Lokru L Haltarve I
Lummelu Etebols I|
Lärbro Grauns Klint I
Lärbro Grauns Klint I
När Hörte I (block)
Rute Bunn I
Stenkuml Martille I
Viklau L Tune I
Västerh Suderbys I
Väster Suderbys II
Väster Suderbys III
Västerh Suderbys IV
Most grinding grooves blocks are also polished on other surfaces. The whole block's upper side is usually more or less polished. Grinding can be performed before the advent of the grooves or after. - It can often be determined which one did come first. Grinding has usually first and foremost been done as a flat grinding of the upper level, but also there are different types of grinding hollows which, may be difficult to distinguish from certain types of shallow and wide grinding grooves. In some cases there is very short and wide grinding depressions with ribbed pattern so is the stone Findarve I from Rone (Raä Burs 179). See Fig 27. |
But there are also polished blocks without any grinding grooves at all. They occur in the same area as the grinding grooves stones and they therefore probably belong to the same period. The National Heritage Board's archaeological inventory has been included. However, they seem generally only to be found on southern Gotland. It is probably so that they can only have been highlighted there. Then there can be questioned if a certain block should be counted, when it can be difficult to determine if there is a real grinding surface or not.
Figure 27. Findarve I. Note the grinding surface below the two grinding grooves to the right.
Here are acc. National Heritage Board's list:|
Sorted on parishes
(The numbers indicate the number of blocks).
The grinding grooves that are located on or near the original site have been included. For the moved blocks whose locality is known, the locality is indicated. Moved blocks whose locality is not known, however, has not been included. The map attempts thus to show the original distribution of the grinding grooves. Moving frequency of grinding grooves blocks have not been spread evenly over Gotland. For example, close to some ports many blocks have been lost in the construction of the harbour. Around Visby is the plot of the grinding grooves stones, which is reportedly due to the great port construction (mostly in the late 1800's, but the block has travelled to the port from further away located parishes during the last port construction).|
The maps attempt thus to provide a picture of the original distribution of the grinding grooves on Gotland. It should be noted that many grinding grooves (especially on southern Gotland) is at such levels that they almost certainly are honed on the former shoreline. Grinding grooves groups are also at levels several meters above the lowest. Probably beach levels, for the time grinding grooves honed in, moved up and down as much as that. Approximately 70% of all grinding grooves are above the 10-meters level and probably was at no time touched by the sea. When visiting these, one is struck by the fact that there former should have been water (former lakes). - Before the big drainage was about 12% of Gotland covered by water. Some of the grinding grooves, however, are on top of the hills quite high over the past water. - The grinding grooves placements are just in the locations where they had a clear view!
The following pages present a list of all known Gotland grinding grooves. The list is based primarily on the National Heritage Board's ancient memory inventory. The grinding grooves are presented by parishes in alphabetical order (sockenvis). Each block or rock has been named after the property designation with Roman numerals in the order (almost) as they have been recorded in the said inventory, or if they are not included in the order in which they have come to the author's knowledge. Block or rock individual names can be found in the leftmost column. The second column is the National Heritage Board inventory (Raä) number (if available). The third column is the kind of stone according stone inventory, indicated with an abbreviation. (In some cases the author has corrected the list).|
The following abbreviations are used: g = granite, rgr = red granite, vgr = white granite, sgr = black granite, ggr = gray granite, lgr = light granite, mgr = dark gray granite, gn = gneiss, di = diorite, ks = limestone, ss = sandstone, rss = red sandstone, kv = quartz, kvs = quartzite sandstone, rpb = reddish porig rock, mb = dark rock, fr = fragmentary, fgr = fragmentary granite, frgr = fragmented red granite, fks = fragmental limestone, fss = fragmental sandstone, FH = solid rock, FS = gone.
The fourth column is the number of grinding grooves on the block or rock.
Finally, local location is specified in latitude and longitude, and / or distance from a benchmark.
A lot of grinding grooves compared to the National Heritage Board Antiquities inventory has been added during this work. Often more blocks with grinding grooves have been found on old buildings and more grinding grooves in solid rock has been discovered adjacent to previously known. Also blocks outside of the National Heritage Board designated premises have come to the author's knowledge. A number of blocks or rocks should further still be detectable and the author hopes to be aware of these. One reason that this list is published is that people out in the districts could then compare this with the grinding grooves that they know.
Another reason for publishing the list is that people with home care ambitions is to have control of the grinding grooves located in the district and to notice if any blocks has been removed from the site. Grinding grooves are protection under Antiquities Act and it is very important that the blocks will not move out of position.
The author has repeatedly noticed that blocks have been removed since the last census (conducted in the 1970's).
Created by Sören Gannholm
To first page
Page last modified