What is Whole Horizon Analysis?
Is there any archaeological evidence for astronomical use of these views?
What about trees? (Wasn't the land forested then?)
What is the difference between then and now?
What about stars and planets?
How accurately were the monuments sited?
Why are there so many monuments?
Why are some sites surveyed more thoroughly than others?
How accurate are the pictures?
Why is this research important?
What do the Data Tables mean?
What is Whole Horizon Analysis?
Whole Horizon Analysis literally means measuring the profile of the Whole Horizon, thus enabling a search for consistent patterns of correlation between horizon markers and astronomical events. This is in sharp contrast to the more traditional archaeoastronomical technique of only measuring monument axes, which is based on the comforting but incorrect premise that only if something man-made is pointing at it can it be shown to be deliberate (See: Is there any archaeological evidence for astronomical use of these views?). Whole Horizon Analysis is the only methodology to date which truly complies with Hawkins' criteria five (1973:290) "All Possible Alignments at a Site Must Be Considered".
While other techniques could be used and may be prove to be preferable, this research has been done with theodolite and camera. Survey of the horizon in fixed azimuthal increments was, after trial, rejected. Small increments are too time consuming and larger ones only produce an approximation of the true profile. The method adopted has been the survey of identifiable "break points" such as high points, low points, steps and other prominent features. Photographs were used to record horizon profile, relationship of monument to horizon, position of instrument etc.
In practice, survey was largely concentrated on the zone occupied by the zodiac rather than the whole horizon. The more northerly and southerly zones were checked periodically without any consistently recurring declinations being obvious though some approximations to due north-south were spotted, both as horizon points and axes of monument components. However, recent developments in the use of drafting software to accurately map photographs of the horizon onto the survey data have lead to a realisation that the marking of north and south points was actually very important so the survey process has been modified again.
Reference: HAWKINS, GERALD S. 1973. Beyond Stonehenge. London: Hutchinson.
Is there any archaeological evidence for astronomical use of these views?
To summarise the standard argument / premise: If a "significant" astronomical event is marked by a natural feature, then this may only be regarded as intentional should something man-made point at it as well.
This viewpoint is derived from Hawkins' criteria two (1973:289), "Alignments Should Be Restricted to Man-made Markers". It is demonstrably incorrect. In the case of a single monument any coincidence between axes of monument features and celestial events on the horizon can not be demonstrated to be anything more than a coincidence. To demonstrate deliberate intent, one must be able to show repeatability and consistency. This study demonstrates repeatable and consistent patterns of coincidence between horizon features and celestial events in both axial and non-axial directions. It also reveals why standard studies of monument axes have failed to show such patterns:
- Monument axes are context sensitive: They mean different things in different circumstances and understanding the horizon provides the context.
- Axes are generally similar to hand signals in that they vary from the waved "look at that over there" for something obvious, to the "look directly along this exact line" for something easy to miss or sometimes even a void in the sequence.
- Unusually extreme axes appear to be a convention meaning "pay particular attention to horizon quadrants perpendicular to the normally preferred axis".
- Standard studies have made assumptions about which declinations are significant and anything not clearly indicating one of those has been written off as not significant. This study similarly started off looking for evidence for or against the observation of lunar extremes or Thom's solar calendar but these basic assumptions have been modified by the survey data to include other declinations that have been revealed both by their axial indication and their repeated recurrence at other key points on the horizon.
- Understanding the horizon also reveals the existence of other, more subtle, indicators within monument structures, which is yet more evidence of deliberate intent.
To summarise the archaeological evidence that these sight lines really were used:
- The monuments are where they are and if they were somewhere else then their relationship with the horizon would be both different and less astronomically significant. There are many examples in the catalogue that demonstrate the importance of exact position.
- Monument axes can have various meanings which are context sensitive, as discussed above. Thus to properly understand the meanings of various architectural features it is necessary to be aware of the relationship of the horizon and sky as seen from the position occupied by the monument.
- There is some evidence, from excavations, of features that are consistent with the setting out arrangements necessary for correctly positioning the monument in relation to the horizon.
References:
HAWKINS, GERALD S. 1973 Beyond Stonehenge. London: Hutchinson.
THOM, A. 1967 Megalithic Sites in Britain. Oxford: Oxford University Press.
What about trees? (Wasn't the land forested then?)
Trees and buildings are certainly a problem now but the prehistoric problem resolves itself into three parts:
- There were no trees obscuring the view of the horizon from the monument at the time it was built.
- The correlation between horizon features and astronomical events found at every site is too consistent for it to be a chance occurrence, therefore the monument builders could see the horizon.
- Trees on distant horizons subtend such a small angle that they have no significant effect and therefore their presence or absence is inconsequential.
- The presence or absence of trees on local and middle-distance horizons would be significant and measurable.
- If the horizon had been tree covered and not used astronomically it is unlikely that significant correlations would be found now.
- If such a horizon had been completely tree free when the monument was built then the astronomical correlations found now would be consistently accurate. This is the usual case.
- If the horizon had been used calendrically but had been tree covered when the monument was built, then a naked modern horizon would have a lower altitude.
Thus there would be a regular shift in the measured declinations that would allow the height of the tree canopy to be estimated.
Close inspection of measured declinations is necessary.
A very few sites have middle-distance horizons where this may be the case.
More often there are small variations that might indicate the presence of scrub woodland in lower parts while higher ground remained tree free.
Generally speaking the monuments do not occur in places where it would be necessary to clear a very large area in order to see the horizon. There is also a noticeable difference in the type of location favoured by monuments from different periods that is attributable to there being heavy tree cover in earlier periods and considerable clearance in later times. Remember that all these surveys show horizons that have been pre-selected for their suitability by the monument builders and such horizons are not the general case, as one learns to recognise whilst travelling between them.
What is the difference between then and now?
The Obliquity of the Ecliptic varies cyclically over 25,800 years approximately between 24.3 and 21.9 degrees and has been falling from the maximum value for some thousands of years. In 1800BC, obliquity was about 23.9°, reducing during the Bronze Age by roughly 0.01° per 100 years. Current obliquity is about half a degree lower than it was during the Bronze Age and as the solar disc is about half a degree across it follows that for practical purposes, at the solstices, the sun now falls short of its Bronze Age rise / set declinations by its own diameter. The equinox is the same now as ever of course and the whole central part of the year is similar enough. By the cross-quarters, a Bronze Age horizon calendar is about one day off the modern equivalents but our solstices occur at points on the horizon that would have been about 10 days away from the Bronze Age ones.
Thus: In the Irish Neolithic, obliquity was slowly reducing after a long period of loitering around its maximum value but now we are accelerating towards the mean value.
In practice this means: The tilt of the earth's axis is a bit less now than it was then, so the difference between the extreme positions is also a bit less. For the moon this means that the whole range of lunistice positions is shifted towards the centre by the same amount as the solstices.
- Summer Solstice Sunrises: The Sun Then lifted off the horizon at about the same position it first becomes visible now.
- Winter Solstice Sunrises: The Sun Now lifts off the horizon at about the same position it first became visible then.
- Sets are the mirror image.
There is no point in locating these positions by compass except as an aid to having an idea of where to look. Assessing the axis of a monument with a compass and then claiming it to be accurately aligned to some solar or lunar event is simply not credible:
- A compass is not accurate enough.
- Horizon altitude makes a significant difference that is very difficult to estimate.
- Real horizons are rarely horizontal.
A useful rule of thumb is to use a thin stick (biro) held at arms length (ish) to estimate diameters along the horizon.
CAUTION: NEVER LOOK AT THE SUN WITHOUT PROPER EYE PROTECTION.
What about stars and planets?
Alexander Thom claimed alignments with certain stars and others have suggested certain planetary alignments, notably venus.
I always considered it inadvisable to seek stellar explanations for declinations within the luni-solar zone until both the solar and lunar situation was fully understood. Events have (seemingly) proved me correct in this, in that I have found no recurring declinations that are not best explained by my suggested megalithic luni-solar calendar. Even so, I have checked for patterns of possible stellar declinations and they have not appeared to date. A possible exception to this is the occasional recurrence of declinations around the low 30's which could indicate an interest in indicating the approximate zodiac limits. Wondering if they might have used particular stars to set up a site more quickly than could be done by waiting for the moon, I looked for evidence of that but failed to find any.
Any experienced observer has no need to mark the rising or setting positions of stars. Stellar rise/set points are constant for any given location within the scale of a human lifetime even though they do change significantly over long time periods due to precession. One recognises a star not only by its brightness and colour but, perhaps more so, by its relationship with others. Therefore, knowing the patterns in the sky, one knows when & where a star will set because it is obvious and where & when one will rise by what is already visible in the sky.
A lot of people get hung up on the fact that the early Egyptians created alignments to observe the heliacal rising of Sirius but it must be understood that they created these alignments to measure first appearance at a specific altitude so as to be as time-of-year specific as possible. It should also be understood that Sirius was a significant indicator of seasonal fertility cycles, at that time and at those latitudes but less so outside those parameters. There is not the slightest indication that megalithic astronomers were at all interested in absolute altitudes. They manipulated the shape of the horizon so that its changes measured cyclical sky patterns.
As for the planets, their declinations do indeed vary. However, I looked at the patterns and have seen no useful purpose to be gained by marking any of their extreme positions. Again, the survey data shows no recurring declinations not best explained by my suggested megalithic luni-solar calendar.
How accurately were the monuments sited?
Almost certainly to within 10m of the optimal position and quite possibly better than that. As it's all about angles, horizon distances and working compromises it is difficult to be specific. To generalise however, the survey results indicate that:
- Monuments 200m or more apart were using a broadly different set of correlations, a different "picture".
- Monuments within 100m or so of each other were using a different set of compromises within broadly the same set of correlations, a different "focus".
- Monuments within 10m of each other may generally be regarded as being in the same place. This would seem to occur when options 1 & 2 were not available to the builders of the later monument. However, even smaller separations may be critical in at least one direction, as at Reenmeen West.
Why are there so many monuments?
There are even fewer now than there were, many have been destroyed over the years.
Whatever else their social function, they were clearly intended as bases for regular celestial observations. To suggest that this is not so is to imply the existence of another set of completely unmarked horizon calendars in order for them to have the knowledge and skill required to site the monuments in the way that they have.
The monuments often occur in local groups and the reasons for it would appear to be as follows:
- On first colonising an area people would need to set something up fairly quickly for the purpose of calendar regulation. This first monument would often be less than perfect. It might only have partial functionality due to limitations of the local landscape, or it might be necessary to set up near to someone else's pre-existing site.
- As soon as resources (or land clearance) permitted, a second monument would be built in the same general area. This one would be sited so that it complemented the functionality of the first.
- As territory expanded, new monuments would be built as suitable sites presented themselves until the combined functionality was sufficient. Building might then cease in one area and the process would recommence some distance away in a new district.
- The object was to achieve accurate cover of all parts of both the solar year and the lunar nodal cycle. A certain amount of redundancy was also desirable because then observers from different sites could compare results and sometimes, due to weather conditions, one would achieve a successful observation when another did not.
- Apprenticeship might well be another factor. The skills would have to be passed on by a training process. This would probably involve memorising a lot of information but would also, of necessity, mean attending the observations of a "master" for at least half of an 18.6 year lunar node cycle. The apprentice, in turn, would design and see to the building of his own contribution to the improvement of the network of observing stations.
Why are some sites surveyed more thoroughly than others?
- Early surveys were much less comprehensive than later ones. It took some time to fully realise the necessity of thoroughly checking all four horizon quadrants rather than just doing a quick scan in the axial directions.
- Some sites have still not been resurveyed with that knowledge and these serve, to some degree, to indicate the nature of the development of the surveying regime.
- Some later surveys may be only partially completed. Perhaps an initial survey was curtailed due to lack of time or poor weather and a further opportunity has not occurred.
How accurate are the pictures?
Pretty good and better than anything else currently available. See Technical Notes on the Pictures.
Why is this research important?
The implications of this research are important in three main areas:
- Understanding Human Development.
- These results demonstrate that humans were a lot cleverer a lot earlier than we have previously been prepared to admit. They also reveal a different way of perceiving the physical world which is based on being constantly aware of the shape of the landscape and how this may be manipulated by changing the observer's position.
- Archaeological & Archaeoastronomical Methodology.
- I would argue that Whole Horizon Analysis should be developed as a standard methodology and that it should be widely applied until we are certain which site types were NOT astronomically sited. The implications for archaeological excavations are that traces of site location and setting-out works could occur in the ground around monuments, perhaps as much as 100m away from them. Because monuments were sited and laid out astronomically, excavation grids should be accurately surveyed and their relationship to True North established to as high a precision as possible.
- Heritage Protection & Planning Regulations.
- It is now clear that the view of the horizon is as much a part of the monument as the stones from which it is constructed. It is also probable that the stone monument is the centre of a zone of activity that extends some considerable distance from it. Wherever possible the existing views should be preserved and buildings that cut the horizon should not be permitted. If they are, then a full Whole Horizon Analysis should be done first. Forestry and land reclamation works should be discouraged from going right up to the edges of monuments as they commonly do at present.
What do the Data Tables mean?
The data table supplied for each site contains a summary of the survey data for that site and is divided into eight columns:
- The first column contains a brief descriptive name of a horizon point. These names are intended to be sufficient to identify the point again either in reality or on a photograph. To avoid confusion with multiple directions, the two sides of a horizon feature were referred to as the north and south sides irrespective of actual orientation. More recently the terms left and right have been used instead. The terms approx. and ish mean that the absolute accuracy of the surveyed point is uncertain (though generally only to a small amount) - this is frequently due to vegetation but may have other causes. A commonly used abbreviation is "c" for centre e.g. Hilltop c ish, another is "B4" for before.
- Az: Azimuth of the point, measured clockwise from true north as determined by timed sunsights.
- Alt: Altitude (Elevation) of the point. The measured angle above or below the horizontal plane.
- AppDec: Apparent Declination of the point, calculated from Az & Alt and corrected for refraction using actual atmospheric temperature & pressure readings.
- LunarDec: Apparent Declination corrected for Mean Lunar Horizontal Parallax.
- Event: The identifier for a calendrical/celestial event.
- EventDec: The declination of the centre of the solar/lunar disc for the named calendrical/celestial event. NOTE that, unless otherwise stated, EventDecs are calculated for about 1800BC and an obliquity of the ecliptic of 23.91°. In 3000BC this would have been about 0.1° more and by 800BC about 0.1° less.
- Error: The difference between the declination of the point and the EventDec. Because the mean diameter of both sun and moon is 0.52°, an "error" of +0.26° is an accurate upper limb and an "error" of -0.26° is an accurate lower limb.