Magure Zimbabwe Site, Distrito de Barue, Província  de Manica Moçambique

 This site was first identified by H.A Wieschhoff in his book, The Zimbabwe-Monomotapa in South-East Africa, George Banta Publishing Company, Menasha Wisconsin, USA, 1941. Wieschhoff spent a considerable time examining both Zimbabwe and Nyanga Sites in the Nyanga District of Zimbabwe and the Barue District of Moçambique.  Most of his time in Moçambique was sped at the Niamara Site high up on Serra Mulanda about 16 km north west of Magure. (But see below) He spent only 5 days at Magure mapping the site and making two test trenches. His account of Magure is reproduced below as well as an amended map with comments as the result of a recent site visit made on Saturday the 10th of March which is also reproduced below.

The access too the site is on a small dirt road leading off the main Vanduzi Tete road to the left (West) some 10 km south of Catandica and just north of the Aldeia 3 de Fevereiro. The turn off is signposted «Património Cultural Amuralhado de Magure» The road continues westwards for about 12 km passing though a small trading centre and School, through a farm and then crossing a couple of fast flowing streams, now bridged, thence through several small villages and machambas to a hill just to the east of the Nhacangara River. The UTM Grid reference for the site is  36K 513322E 7993312S and the altitude is 830 m.
    

Wieschhoff wrote in 1941:

“Investigations among the Barwe of this region led to a short examination of another ruin some eight miles southeast of Niamara. These ruins do not lie on top of a high mountain, as does Niamara, but on a rising ground almost in the plains of Mozambique. The natives refer to the little creek in its neighbourhood as Magure, and since there was no other landmark with a native name in the vicinity, this was selected as the name for the site. These ruins must be the most easterly in this section of the country. After I left the Niamara ruins to go to Inyanga, I took part of my group to this site for five days' work. Like Niamara, its existence had never been reported, and it was an absolutely untouched site, but since our time was limited we had to confine ourselves to the mapping of the ruins and to the sinking of a few test trenches (Fig. 28 adapted in plan below).

The difference between the highest and lowest parts of the walls, which follow the contours of the little hill, was nowhere greater than 7.00 m. As at Niamara, the walls consisted of long narrow slabs of "Rhodesia schist" and were in a good state of preservation with very little debris following their courses. They are on the aver-age between 1.20 and 2.00 m high and uniformly 0.80 m thick. The whole ruin is surrounded by a circuit wall, forming an oval, which measures 75 m in its longest extent. A wall, some 50 m in length, bisects the ruins into two parts of almost equal size. The north western section only had an entrance. The natural unevenness of this section had been utilized for the creation of two higher platforms, C and D, and a lower one B.

The highest elevation within the site is the enclosure G, rectangular in shape and built directly upon the natural boulders which project above the soil. The gaps between the boulders, as well as the areas in front of the entrance to this enclosure, were covered with red daga. The size of the chamber as well as the unprepared floor make it seem unlikely that this red daga represents the remains of a daga house, although I am inclined to think that this enclosure served as a platform for granaries, perhaps similar to the kind found in the ruins of Farm Linder (p. 77).

Two test trenches, both 1.00 m wide and 12 and 38 m long respectively, were cut through enclosure A. The first trench was intended to determine the structure of the daga foundations, three of which could be seen. All of these foundations resembled the daga house of enclosure E at Niamara; that is, they were divided into three parts by walls which obviously had been built of wooden poles. The average diameter of these three daga foundations was approximately 6.00 m. unfortunately the first trench yielded nothing except a few sherds and iron arrow points, nor did the additional trench, which was dug parallel to the first in front of one of the foundations, alter the picture. Thus, no objects of any importance were unearthed and the determination of the architectural features remained the only result of this visit.

There are altogether eight hut foundations at Magure: five in enclosure A, one in D, and two smaller ones, only about 4 m in diameter, in C. The foundations examined were discovered resting directly upon the stratum of residual soil.

Comparing the type of stone masonry as well as the hut foundations of Magure with those of Niamara, it becomes likely that both belong to the same period. It is interesting to note that this type of ruin is found not only on top of high mountains but in valleys as well.”

On Saturday the 10th of March Henrik Ellert,  Jonathan Waters (Jono) and myself went up to Báruè to Visit the site and we were agreeably surprised to find that the road was in quite a good condition and were able to drive right to a village about 100 m east of the site in Jono's Landover. Local people were also very helpful in leading us up to the site from the village. The hill on which the  is situated is covered dense high rainfall woodland with abundant ferns and mosses. The main trees that I can remember were pangapanga with some figs amongst the rocks.

The walling is well preserved and the site does not appear to have deteriorated since Wieschhoff visited the site in the 1920s and Roza de Oliveira in the 1960s. Wieschhoff mentions a covering of dhaga on one the walls but we did not notice this on our visit.

Wieschhoff's Plan Adapted

Unfortunately, I forgot to take a copy of Wieschhoff's plan of the ruins. However on examining this on our return it appears to be an accurate plan of the site.  It appear that en
closures A and B are the outer one and D,E,F and G are the inner enclosures.  These enclosures adjoin the WNW sideof the  hill of the hill, which drops sharply some 6m with some small rock overhangs below, perhaps too small to have given any protection from the weather. 

Entrance to inner enclosures

There was only one visible  entrance  to the inner enclosures, though the site is heavily overgrown and a visit in the drier months of the year might reveal others, this was through a narrow defile between boulders.  See above picture

To the right of this and abutting the boulders was what appear to be a semi-circular platform about 1,5m in height It was not part of the general walling and must have had some other function.  

What is it? By stretching the imagination it could be a small conical tower  but I think if was a platform linked to the entrance of the inner enclosure. It is shown in the photo below  

1 The Platform

The standard of the walling is very high, and consists of schist rather than granite, the local rocks. To me appears far more like Zimbabwe tradition type walling than the walling from the Nyanga Terracing culture, which commences some 20 km to the west near the Zimbabwe Border.

There are now no local traditions concerning the site and who built it. In the 1920s a member of the Barue Royal family accompanied Wieschhoff's to Niamara and Magure and he linked the Niamara site to Changamire, Wieschhoff's also thought that the two sites were linked. He said that they were some 8 miles apart that is 13 km. Using the grid references for the two sites  they only appear to be 4 km - as the crow flies - apart. Niamara being NNW of Magure on the western side of the Nhacangara Valley but 400m higher on a spur of Serra Chôa named Serra Mulanda. This probably explains the different access used on visited to the sites. Niamara has always been accessed from Serra Chôa, whilst Magure was accessed from the route described above from the plains below.

The  UTM Coordinate the sites are

1. Magure    36 K 5133 E 79933 S
2. Niamara  36 K 5106 E 79960 S

    

The site is in Rotanda not Cheringoma. (not sure how that came in!)

Messambudzi (Rotanda). This site lies on the Chinhadembué Range just north of Messambudzi in the Rotanda administrative post in Sussundenga not far from the Zimbabwe border. The site is very overgrown but the walls and platform are in a good state of preservation. Oliveira the Portuguese archaeologist who visited the site in the 1960s thought that there was a ‘conical tower’ in the middle of the site but my ‘gut feeling’ was that this was more like a platform. We were only able to stay at the site for a few minutes, as we had to get down the mountain before dark. The local ruling lineage is Mucimua, the régulo is in the picture, his people belong the shumba clan and related to Mutasa migrated to the area in about the 18th century from Manica. They attributed the ruins to the Changamire Rozvi but say when they arrived in the area they found no one there (unlikely as their neighbours the Gutsa shava lineage who arrived just before them say the found a people called the matsonno who appear to have been the pre Rozvi Tewe (Quiteve) rulers in the area.) The Rozvi migrations into Tewe and Sanga occurred at the end of the 17th century. I think the site predates this and probably goes back to the Zimbabwe culture itself. I will paste in some more information about some neighbouring’s sites shortly.

To Start the blog...

Eastern Shona Origins and linguistic classifications.

It is now accepted following numerous archaeological investigations that the Zimbabwe culture developed from the Gumanye Culture, which in turn developed from and was a successor to the Mupungubwe /K2 culture based on the upper Limpopo Valley near to the present borders of Zimbabwe, Botswana and South Africa.
From Great Zimbabwe itself the culture spread out far and wide, to the Zambezi Valley, the Zimbabwean Plateau, Moçambque between the Zambeze and the Limpopo and finally to the most northern parts of South Africa. The expansions of the culture was associated with an expansion of Shona-speaking people, or perhaps Shona-speaking ruling lineages. This expansion may, or may not, have resulted in language change.
Before any arguments are developed we should note that a change in archaeological assemblages do not necessarily mean a change in language, or for that matter the physical population.
Prior to the 1980s, the Shona language, or the Shona group of dialects were generally classified on their own and placed in an intermediate group between the Eastern Bantu Group — which includes Nyanja/Chewa and Sena — and the South East Bantu group — Shona Venda, Tsonga, Nguni and Sotho/Tswana.
However, since the early 1980s, and perhaps concurrently with the increasing archaeological work and the discoveries linking the origins of the Zimbabwe / Gumanye Culture in the south, the Shona language itself has been placed by most authorities in the South (East) Bantu Group. This I query, at this stage more on empirical and anecdotal grounds rather than on qualified research.
Linguistic Affinities Shona and its Dialects
Shona is essentially a group of closely related dialects, thee principal ones being;
• Karanga in the Masvingo Province of Zimbabwe;
• Zezezuru in the Central parts of Mashonaland in Zimbabwe;
• Korekore and Tavara in northern Zimbabwe and adjoining parts of Moçambique;
• Manyika in eastern Zimbabwe and neighbouring parts of Moçambique and here including some Báruè;
• Tewe (Teve) in Moçambique;
• Ndau in Moçambique and South East Zimbabwe;
• And finally Kalanga in South West Zimbabwe and neighbouring parts of Botswana.
Ndau and Kalanga differ more from standard Shona than the other dialects and may have, have had, links with other languages.
Standard Shona was developed in the colonial period and was based principally on the Karanga and Zezezuru dialects, but it is now spoken and written throughout Zimbabwe and used on the Zimbabwe Broadcasting Corporation. In Moçambique the Shona dialects have retained more of their identity and Báruè (Barwe), Manica (Manyika), Tewe and Ndau are recognised as separate, if related, languages and are all used in broadcasts in Radio Moçambique's regional services. Shona orthography differs in Moçambique as the Portuguese alphabet is employed as opposed to the English but this is really the result of separate colonial experiences. .
The neighbouring languages to Shona in the north and east, all belong to the Eastern Bantu linguistic group are;
• The various Zambian languages to the north of the Zambeze;
• Senga in the Zumbo area;
• Nyanja/Chewa in the area to the north of the Zambeze between Zumbo and Tete;
• Nhungwe in the Tete Changara Area;
• Sena in the Tambara Chemba Areas;
• And Sena/Tonga/Bangue
South between Beira and the Rio Save Shona (Ndau or Sanga) is spoken along the coast and even south of the river towards Vilancoulos. The Boundaries between Shona and the Southern Bantu group of languages are
• Tsonga (here the Tswa and Hlengwe dialects) roughly along the Save and Runde Rivers (but not always), and then goes across country to the Bubye and Limpopo Rivers.
• A small number of Venda-speaking people are found to the north of the Limpopo and Shona (Kalanga) is spoken in NE Botswana. [Perhaps a map should be inserted here].
These linguistic boundaries have not always been static. There is for example evidence than Shona was far more widely spoken in the southern east of the area prior to the Hlengwe advance in the late eighteenth century. In these areas the boundaries between Shona and some neighbouring languages are very well defined, and can be very sudden. One village may speak Shona and the next another language. This is especially the case in the southwest along the Shona / Tsonga divide and is clearly the result of a comparatively recent migration of the Hlengwe and Tswa into what was formally Shona-speaking country. The two languages are not mutually comprehendible.
However, in the north and east Shona appears to blend into other languages far more gradually. It appears to me that this is generally where Shona dialects come into contact with eastern Bantu languages, that is Chewa/Nyanja, Nhungwe, and Sena/Tonga, especially the latter. The Báruè language is generally recognised as a Shona-related dialect, indeed, in the west of the area where it is spoken it is very close to Manica, but in the east and north Sena / Tonga and Nhungwe become more evident. It would appear to me that this indicates a long relationship between eastern Shona and these languages.
I have asked many central and eastern Shona-speakers, which languages it is easier for them to understand, (assuming that they have no prior knowledge of these languages), the Nyanja Sena languages, on the one hand, and the southeast Bantu Group of languages on the other hand. In all cases they have replied that the Nyanja Sena group are easier to understand. A number of very common words that are used in Shona and the northern languages, for example ‘iwe’, are not found as far as I am aware in the southeast group. [More examples needed here]
I suspect that Shona is more an intermediate language between the Eastern Bantu Group of languages and the South (Eastern) Group of languages and that the former classification reflected a truer picture of the actual relationship between Shona and the eastern and south Bantu linguistic groups than the current one. Of course all the languages discussed are fairly closely related, nevertheless the actual relationships between the languages can help to solve both historical and archaeological problems.
Eastern Shona Origins and linguistic classifications.
It is now accepted following numerous archaeological investigations that the Zimbabwe culture developed from the Gumanye Culture, which in turn developed from and was a successor to the Mupungubwe /K2 culture based on the upper Limpopo Valley near to the present borders of Zimbabwe, Botswana and South Africa.
From Great Zimbabwe itself the culture spread out far and wide, to the Zambezi Valley, the Zimbabwean Plateau, Moçambque between the Zambeze and the Limpopo and finally to the most northern parts of South Africa. The expansions of the culture was associated with an expansion of Shona-speaking people, or perhaps Shona-speaking ruling lineages. This expansion may, or may not, have resulted in language change.
Before any arguments are developed we should note that a change in archaeological assemblages do not necessarily mean a change in language, or for that matter the physical population.
Prior to the 1980s, the Shona language, or the Shona group of dialects were generally classified on their own and placed in an intermediate group between the Eastern Bantu Group — which includes Nyanja/Chewa and Sena — and the South East Bantu group — Shona Venda, Tsonga, Nguni and Sotho/Tswana.
However, since the early 1980s, and perhaps concurrently with the increasing archaeological work and the discoveries linking the origins of the Zimbabwe / Gumanye Culture in the south, the Shona language itself has been placed by most authorities in the South (East) Bantu Group. This I query, at this stage more on empirical and anecdotal grounds rather than on qualified research.
Linguistic Affinities Shona and its Dialects
Shona is essentially a group of closely related dialects, thee principal ones being;
• Karanga in the Masvingo Province of Zimbabwe;
• Zezezuru in the Central parts of Mashonaland in Zimbabwe;
• Korekore and Tavara in northern Zimbabwe and adjoining parts of Moçambique;
• Manyika in eastern Zimbabwe and neighbouring parts of Moçambique and here including some Báruè;
• Tewe (Teve) in Moçambique;
• Ndau in Moçambique and South East Zimbabwe;
• And finally Kalanga in South West Zimbabwe and neighbouring parts of Botswana.
Ndau and Kalanga differ more from standard Shona than the other dialects and may have, have had, links with other languages.
Standard Shona was developed in the colonial period and was based principally on the Karanga and Zezezuru dialects, but it is now spoken and written throughout Zimbabwe and used on the Zimbabwe Broadcasting Corporation. In Moçambique the Shona dialects have retained more of their identity and Báruè (Barwe), Manica (Manyika), Tewe and Ndau are recognised as separate, if related, languages and are all used in broadcasts in Radio Moçambique's regional services. Shona orthography differs in Moçambique as the Portuguese alphabet is employed as opposed to the English but this is really the result of separate colonial experiences. .
The neighbouring languages to Shona in the north and east, all belong to the Eastern Bantu linguistic group are;
• The various Zambian languages to the north of the Zambeze;
• Senga in the Zumbo area;
• Nyanja/Chewa in the area to the north of the Zambeze between Zumbo and Tete;
• Nhungwe in the Tete Changara Area;
• Sena in the Tambara Chemba Areas;
• And Sena/Tonga/Bangue
South between Beira and the Rio Save Shona (Ndau or Sanga) is spoken along the coast and even south of the river towards Vilancoulos. The Boundaries between Shona and the Southern Bantu group of languages are
• Tsonga (here the Tswa and Hlengwe dialects) roughly along the Save and Runde Rivers (but not always), and then goes across country to the Bubye and Limpopo Rivers.
• A small number of Venda-speaking people are found to the north of the Limpopo and Shona (Kalanga) is spoken in NE Botswana. [Perhaps a map should be inserted here].
These linguistic boundaries have not always been static. There is for example evidence than Shona was far more widely spoken in the southern east of the area prior to the Hlengwe advance in the late eighteenth century. In these areas the boundaries between Shona and some neighbouring languages are very well defined, and can be very sudden. One village may speak Shona and the next another language. This is especially the case in the southwest along the Shona / Tsonga divide and is clearly the result of a comparatively recent migration of the Hlengwe and Tswa into what was formally Shona-speaking country. The two languages are not mutually comprehendible.
However, in the north and east Shona appears to blend into other languages far more gradually. It appears to me that this is generally where Shona dialects come into contact with eastern Bantu languages, that is Chewa/Nyanja, Nhungwe, and Sena/Tonga, especially the latter. The Báruè language is generally recognised as a Shona-related dialect, indeed, in the west of the area where it is spoken it is very close to Manica, but in the east and north Sena / Tonga and Nhungwe become more evident. It would appear to me that this indicates a long relationship between eastern Shona and these languages.
I have asked many central and eastern Shona-speakers, which languages it is easier for them to understand, (assuming that they have no prior knowledge of these languages), the Nyanja Sena languages, on the one hand, and the southeast Bantu Group of languages on the other hand. In all cases they have replied that the Nyanja Sena group are easier to understand. A number of very common words that are used in Shona and the northern languages, for example ‘iwe’, are not found as far as I am aware in the southeast group. [More examples needed here]
I suspect that Shona is more an intermediate language between the Eastern Bantu Group of languages and the South (Eastern) Group of languages and that the former classification reflected a truer picture of the actual relationship between Shona and the eastern and south Bantu linguistic groups than the current one. Of course all the languages discussed are fairly closely related, nevertheless the actual relationships between the languages can help to solve both historical and archaeological problems.


Any comments?