Displaying astronomical alignments in academic papers

Jul 6, 2006 Alun Science, The Past 1 Comment Archaeoastronomy, Archaeology, Astronomy, Georg Zotti, Mapping

[Cross-posted to i-Science]

Astro­nom­ical align­ment at Segesta

There’s a couple of paper which have come out recently which use dif­fer­ent tech­niques for indic­at­ing astro­nom­ical align­ments at archaelo­gical sites. The image above is one I put together for a poster to show why hori­zon alti­tude is import­ant as well as azi­muth. It’s quite tight, so it’d be no good if you wanted to see where sun­rise was in mid­sum­mer for instance, and chart­ing the paths of astro­nom­ical bod­ies over a site is a prob­lem. By and large you can treat a site as a small flat area, so there’s not usu­ally any car­to­graphic prob­lems in account­ing for the curvature of the earth. The sky in con­trast is very curved over every archae­olo­gical site, so how to you dis­play that in a paper?

The Mega­lithic Portal put me on to an inter­est­ing art­icle pub­lished in Inform­a­tion Visu­al­iz­a­tion: A Sky Dome visu­al­isa­tion for iden­ti­fic­a­tion of astro­nom­ical ori­ent­a­tions by Georg Zotti. The abstract includes:

This paper presents a novel dia­gram com­bin­ing archae­olo­gical maps with a folded-apart, flattened view of the whole sky, show­ing the local hori­zon and the daily paths of the Sun, Moon and brighter stars. By use of this dia­gram, inter­est­ing group­ings of astro­nom­ical ori­ent­a­tion dir­ec­tions, for example, to cer­tain sun­rise and sun­set points could be iden­ti­fied, which were evid­ently used to mark cer­tain days of the year.

Unfor­tu­nately Inform­a­tion Visu­al­iz­a­tion isn’t a journal archae­olo­gists get, and it costs $30 to down­load the paper. What I can talk about though is a con­fer­ence paper on his own site: A Sky Dome Visu­al­isa­tion for Iden­ti­fic­a­tion of Astro­nom­ical Ori­ent­a­tions, which includes in the abstract:

This paper presents a novel dia­gram com­bin­ing archae­olo­gical maps with a folded-apart, flattened view of the whole sky, show­ing the local hori­zon and the daily paths of sun, moon and brighter stars. By use of this dia­gram, inter­est­ing group­ings of astro­nom­ical ori­ent­a­tion dir­ec­tions, e.g. to cer­tain sun­rise and sun­set points could be iden­ti­fied, which were evid­ently used to mark cer­tain days of the year.

The two look as though they’re likely to be similar.

The idea is actu­ally rather clever and I’ll go through a very sim­pli­fied ver­sion of a dia­gram based on his method.

I’ll start with the blue circle. That’s the hori­zon around the site. Any­thing on that blue line is at a hori­zon alti­tude of zero degrees. The orange circle, that’s the zenith, the point of the sky dir­ectly over­head. That looks weird but in the pro­jec­tion it makes sense. I’ll try to explain.

If you want a sim­ilar example think about the Mer­cator pro­jec­tion of the Earth. The point at the very top of the map, where the North Pole is every­where along that line at the top. The place, which is one point, has been dis­tor­ted in the pro­jec­tion — but that doesn’t mat­ter if we know how to read the map. Sim­il­arly the orange circle marks 90 degrees from the hori­zon. If you look west to the hori­zon and move your gaze up then, on the dia­gram above, you move from nine o’clock on the blue circle along the black spoke to the orange line. When you reach the point dir­ectly over­head you’re on the orange circle. The reason the orange circle goes around the whole dia­gram is that whatever dir­ec­tion you look in, if you move your gaze up you’ll reach that zenith point, so on this dia­gram the zenith has to be vis­ible in all directions.

This is quite a handy way of draw­ing the sky because it’s easy to draw in the hori­zon. For instance if there’s a moun­tain to the south of the site you can shade in the sky as far as the height of the moun­tain. You can also draw paths through the sky. The red circle is a star that just grazes the north­ern hori­zon and cul­min­ates in the south at 45 degrees. It’s a cir­cum­polar star. Other stars have other paths and Zotti’s paper is full of examples.

I’m quite enthu­si­astic about the paper from an astro­nom­ical per­spect­ive. Alas, I’m never going to use the tech­nique, nor is any­one else bar a few people in very spe­cial cir­cum­stances because the tech­nqiue doesn’t work. The green patch in the centre is where you would draw your archae­olo­gical plan, so below is mine.

I’ve put in two temples with par­al­lel align­ments. Or at least I thought they were par­al­lel align­ment when I drew them. If you look to see where they point to, one temple faces 80 degrees and the other 100 degrees. It’s a very simple dia­gram so why wasn’t this spot­ted as a problem?

In his paper Zotti demon­strates the tech­nique with a very dif­fer­ent sort of monu­ment, a cir­cu­lare Neo­lithic enclos­ure. The pre­sumed view­ing point is at the centre of the dia­gram and the align­ments are radial plot­ted from the centre of the plan. Nat­ur­ally his archae­olo­gical plan works won­der­fully. If how­ever you have an archae­olo­gical site with more than one build­ing, or even one with one build­ing which isn’t radial then the pro­jec­tion in the centre of the plan isn’t going to work. To get it to work you’d either need an infin­itely hori­zon circle or else shrink the scale of your archae­olo­gical plan to a point, either of which knock out the whole point of the paper. If the work is going to be of any use it has to be astro­nom­ic­ally sound and archae­olo­gic­ally use­ful. Simply being accept­able in one dis­cip­line isn’t good enough.

It’s a shame because in many ways it’s an eleg­ant sys­tem, but it seems to be totally inad­equate for gen­eral archae­olo­gical use. This is an example of why I’m wary when I hear the word inter­dis­cip­lin­ary. If this is broadly the same as the paper in Inform­a­tion Visu­al­isa­tion then I doubt that they ran it by an archaeologist.

Another per­son tack­ling the prob­lem of dis­play­ing sev­eral astro­nom­ical align­ments in one dia­gram is M Wilson whose Bronze Age astro­nomy: new sur­vey, new hori­zons is the Pro­ject Gal­lery in this issue’s Antiquity. The solu­tion is a lot sim­pler, sig­ni­fic­ant points on the hori­zon are marked. In this case the images look crowded in places, but that’s mainly because of the nature of the Pro­ject Gal­lery where you have to be con­cise and there is a lot of inform­a­tion to convey.

As for the con­tent of the paper, I’m not con­vinced. Again the brev­ity means he doesn’t have the space to fully make his case. My prob­lem is that it appears that points on the hori­zon are arbit­rar­ily chosen to fit sig­ni­fic­ant dates. There are equally inter­est­ing points on the hori­zon that are skipped over unla­belled, so are points being cherry-picked to fit an idea? Also is there archae­olo­gical evid­ence that these sight lines were used? A row of stones in a line as as good a sign as you’re going to get in this period that some­thing inter­est­ing lies in at least one dir­ec­tion, so you can cer­tainly argue for some align­ments, but what of the unmarked align­ments? Frus­trat­ingly eth­no­graphic evid­ence sug­gests a lot of astro­nom­ical activ­ity will leave no archae­olo­gical trace. It’s not whether or not sky obser­va­tion occurred, it’s how much you can verify.

What is impress­ive is that M Wilson is put­ting his work in front of the eyes of archae­olo­gists on the site of the UK’s top archae­ology journal. There’s a lack of affil­i­ation, so I don’t know if he’s an ama­teur or not, but if he is then he’s going exactly the right way of get­ting the atten­tion of archaeologists.