General Introduction

In 1982 I accepted a NEH grant to do an edition and translation into English of Volume A of Giovanni Antonio Cavazzi’s “Missione evangelica nel Regno de Congo” an Italian description of Central Africa completed around 1668. This text is widely known as the “Araldi Manuscript” since it was owned by the Araldi Family of Modena, Italy, and it had first been described by Giuseppe Pistoni in 1969. I managed to complete a transcription of the manuscript, and thanks to the assistance of Maria Luisa Martini and particularly Carolyn Beckingham, a full translation. I also wrote an introduction and annotated the whole text. In this way, I had completed, successfully, all the requirements of the NEH grant.

Of course, I recognized that the next logical step was to publish the results of my work, and here I found myself balking. I realized that between the three of us, there were problems, gaps and difficulties with the translation, and as I began working through the text again in 1988 this became even more clear. Other material and work intervened and gradually the final production of the Araldi Manuscript moved farther and farther from my thoughts, though it always remained in the back of my mind, and perpetually in New Year’s resolutions.

My somewhat guilty complacency, however, was broken in August, 2001.  On the 12th of August, I received an e-mail message from  Paul Hair, the well known editor of African history related texts.  Paul was one of the first scholars to make extensive use of e-mail, and we had corresponded with it for some years. Paul’s message said, in effect, “this is the last e-mail I will write to you, please publish the Araldi Manuscript.” The very next day, I learned that Paul had died and that this message carried the moral equivalent of a last request.

At that point I felt that I was now compelled to publish my efforts, if nothing else as a memorial to my late friend.  However, even with this encouragement I found myself reluctant to take the last step, and once again procrastinated on publishing.  Seven years went by, and finally in 2008, I decided that I would publish the Araldi MS on the web, in a new experimental format, of which I hope that Paul, always interested in new technology and its application in scholarship, would approve.

I decided that since I wanted the text to have feedback, I would mount it on a blog, and in August and September of 2008, I put up the whole text on a Google blog site, with the title “Central African History.”  While the blog publication was available to all on the web, I could assuage my guilty conscience, but I also realized that it was not fully adequate.  Visitors to the blog could rapidly see that the system of mounting the text, one chapter of text equaling one blog post, was not very user friendly, and while the comment section allowed my visitors (a total of 2 in 3 years) to make suggestions, corrections or commentary, any further postings would gradually push the text farther and farther down the archive.  So I made no more posts of my own, and finally I have decided to move the text to this site.

My original intention was to publish the original Italian manuscript text also, since no one could comment on a text without seeing the original language.  I had digitized the manuscript from a microfilm copy that Professor Joseph C. Miller of the University of Virginia photographed in 1977, and thought I would just mount those.  But here I was concerned, as certainly the Araldi Family of Modena, who owned the copy, would have to give permission for a web publication.  Repeated attempts to reach them failed, however, and I felt that I could not put my digital copy up.  I then considered mounting scanned images of my original transcript of the text, made on a manual typewriter in the summer of 1982, complete with strike-overs, key jams, flying caps and all the other typing errors which word processing put to a permanent end.  I was sure the transcription was my intellectual property.

However, a chance encounter in Angola in 2010 convinced me that the Araldi Family had decided to grant rights to the MS to some Italian scholars, whose names and affiliations I was unable to discover.  This person also told me that they would consider my publishing a transcription to be a violation of their right to make one, and that I might risk a suit should I post it.  Unsure of the law and not willing to be sued in an Italian court, I have decided that until these issues can be resolved, the original language text must remain unpublished by me.

There are additional copies of Miller’s microfilm, one at the University of Virginia, a second at the University of Wisconsin and a third at UCLA which can be borrowed on inter-library loan for interested scholars or other persons, should they wish to check or challenge my work.  In the meanwhile, I hope that visitors to this site can at least have the text of my efforts at translation, as well as my original annotations.

The Araldi Manuscripts

Giovanni Antonio Cavazzi da Montecuccolo’s book Istorica Descrizione de tre regni Congo, Matamba ed Angola , published posthumously in Bologna in 1687, has long been recognized as a standard source for seventeenth century Angolan history and society.  But it was only in 1969 that Giuseppe Pistoni published an account of a hitherto unknown manuscript collection entitled “Missione Evangelica al regno del Congo” attributed to Cavazzi and clearly an earlier draft written between 1665 and 1668.[1] The manuscripts were important, because it had long been known that Cavazzi’s manuscript of Istorica Descrizione had been lost and the published version of the work had been brought to the publisher by Fortunato Alamandini, who acknowledged having edited the material before publishing it.[2] These new manuscripts, owned by the Araldi Family of Modena, which were divided into three thick volumes (which Pistoni designated volumes A, B, C) are clearly ancestral to the longer book, and both Pistoni and Teobaldo Filesi, who also examined the manuscripts, called for deeper study and publication.[3]

Pistoni was able to demonstrate, in his initial study, that the original manuscripts had laid, probably since 1669 or 1670, in the Provincial Archives of the Capuchin order at Bologna until the suppression of the order in the late eighteenth century.  At that time, Michaelangelo Araldi da Modena, a member of the order, managed to have them transferred to his family’s library, where they have remained ever since.[4]

The Araldi manuscripts present few difficulties in identity, since they are autographed in several locations, and both the handwriting and the style of the illustrations match that of other material autographed by Cavazzi and found in the Archives of the Propaganda Fide in Rome.[5]

A Description of the Manuscripts

Any attempt to publish the Araldi manuscripts must deal with their great length of well over 1,500 folio pages.  Fortunately, the useful contents of the material is on considerable fewer pages.  Of these, volume A, entitled, like the other two, “Missione Evangelica” dated in Matamba, 1665, is the most important.  It is divided into three books:  Book 1 is a description of the customs, history and religion of the “Jagas” (Imbangala) of Kasanje and Matamba and runs 145 pages.  Book 2, 224 pages long, is devoted to a long history of the kingdoms of Ndongo and Matamba, but concerned mostly with the life of Queen Njinga, who ruled Ndongo from 1624 until expelled by the Portuguese about 1630 and then in her refounded kingdom in Matamba until her death in 1663.  Book 3 is a short treatise of 41 pages which recounts the life of Kasanje ka Kinguri, ruler of Kasanje, and the unsuccessful attempt by the Capuchins to convert him to Christianity.

Volume B is dated 1666, was completed at the Capuchin hospice in Matamba on 22 August 1665 and deals, after a brief introduction to the history of the Kingdom of Kongo, with the Capuchin mission in Kongo and Angola.  Most of it is composed of biographies of the missionaries, including Cavazzi’s own travels.  Volume C is an incomplete copy of Volume B, dated Luanda, 8 September 1666.  After an examination of this material in 1977, I concluded that for the most part the material on Kongo and the missionary history found in Volumes B and C, while of interest to specialists, was generally not as detailed, accurate or informative as the corresponding sections of Istorica Descrizione.[6] It was for this reason that I decided to concentrate my efforts on Volume A.  The material dealt with in this volume is of the greatest interest to the Africanist, since it is almost entirely about the internal history of the Mbundu people and their customs, while the other volumes deal largely with mission history and only peripherally with the people subject to their activity.

Moreover, while Cavazzi did make a short trip to Kongo in 1664,[7] much of his information on that country was derived from the work of other missionaries, as can be gleaned from the dramatic increase in information in Istorica Descrizione over the brief, and inaccurate, description of Kongo in the Araldi MS B.  A detailed comparison of this new information with known accounts of other missionaries, who spent longer times in Kongo, shows fairly clearly Cavazzi’s debt to their work.[8] This cannot be said, however for Cavazzi’s description of the Mbundu and Imbangala, as Cavazzi spent a dozen years in their country, spoke their language and visited the courts of all the major political units.  There was probably no other missionary in central Africa who would have been better qualified to write on the Mbundu than Cavazzi, and MS A is full of the kind of personal detail indicative of someone writing from intimate knowledge.  Furthermore, it contains some material not found in the corresponding sections of Istorica Descrizione, and even when the material is similar, smaller differences between the two texts help to illuminate questions of interpretation, authority or chronology.  Thus, although Volume B contains some important details and its study should not be neglected, Volume A is clearly the most likely candidate of reasonable size for full scale publication.[9]

Volume A:  Structure and Composition

In addition to the 411 folios divided into three books that make up the body of Volume A, there are 65 pages of front matter, unorganized material drawn from other drafts of Cavazzi’s work, with each page containing numbers from the various numbering schemes of these drafts.  Most of them are consecutively numbered in pencil which run from VIII to XLVI.  The pages in this front matter are most important because they contain the majority of the volume’s vivid color illustrations (18 out of 27), and it was apparently because of the illustrations that they were saved.  Most of the pages either contain an illustration, or are on the back of an illustration, or in the pages immediately surrounding the illustration.  In most cases, the corresponding textual passages can be found in the main body of the text, and they do not contain material that alters the edition of the text.

They do, however, help us to establish more on the date and manner in which Volume A was constructed.  Although the end of the volume indicates that it was completed at Matamba in 1665, there is evidence in the front matter and elsewhere that point to an earlier date of completion.  One need not necessarily trust the dates that Cavazzi placed on the manuscript as being dates of completion, for the frontspiece of Volume B dates it to 1667 while the date written on the dedication suggests it was completed 8 September 1666 and there is no reason to believe that a similar dating system was employed in Volume A.  Cavazzi was not writing for publication, and we do not know exactly what he was thinking in putting on dates.  This is especially true when one considers that the actual text of Book 2 (the life of Queen Njinga) more or less ends in February 1664, although marginal notes update sections of it to 1668.  Cavazzi clearly kept some sort of notes on the period after 1664, because Istorica Descrizione contains a detailed, obviously eye-witness account of subsequent events in Matamba which has no counterpart in MS A or its additions.[10]

Not only was the manuscript apparently completed earlier than February 1664, but Cavazzi must have been rather long in writing it.  The section in Book 2 after page 143 which deals with Queen Njinga’s last days was originally written in the present tense, but was altered by Cavazzi to a past tense, suggesting that he wrote this section before Njinga’s death on 17 December 1663, and went back after her death to revise it.  These indications thus suggest that the present draft of MS A was begun sometime before December 1663 and was completed by February 1664.

However, the front matter demonstrates that the present version of MS A was not the first one, and Cavazzi must have completed at least one, and possibly several drafts of the work before the draft we now possess.  For example, pages XX-XXV of the front matter contain 4 drafts in extremely similar language of the start of the passage on the origin of the kingdom of Ndongo that begins book 2 of the present volume A.  However, all of these show that it was Chapter 1 of Book 3 of its draft, and moreover, the draft on page XXIV includes the last lines of the preceding chapter which deals with missionary lives from a section now found in MS B.  Likewise, on p. LIV, Cavazzi refers to an event which he mentioned in Book 2, Chapter 5, p. 52 in the life of the missionary Buenaventura da Sardegna.  Thus, this earlier draft contained material that was in MS B as well as in MS A of the later draft.  Although the earlier draft or drafts may have been shorter than the entire contents of Volumes A and B in their present form, it must have been quite long, for pp. LVII-LIX of the front matter indicate the presence of a Book 4, chapter 3 in it as well.  This evidence leads us to believe that Cavazzi had done at least one and possibly more drafts of his work before 1663, and that he was constantly reorganizing his work.

It is possible to guess that Cavazzi began work on this text in October 1660 when he reached the court of Queen Njinga in Matamba or shortly after, and his decision to write it might have been his conviction that he was witnessing some sort of miracle in the Queen’s conversion.  Njinga’s conversion had inspired one other Capuchin contemporary of Cavazzi, Antonio da Gaeta, to write a book whose clear purpose was to recount the what he thought as the miraculous events of her conversion.[11] Since Cavazzi came to Matamba to replace him, he may have decided to write such a history of his own, especially since his previous travels in Angola gave him a better background to do a full history of the Capuchin mission than it gave Gaeta.  Although Gaeta had probably finished his book in 1658,[12] Cavazzi seems to have had no direct knowledge of it when he wrote MS A, although he made extensive use of it when he revised his work to do Istorica Descrizione.  Most of Cavazzi’s previous work in Angola had been in Portuguese possessions or those of its close ally, the king of Mpungu a Ndongo.  While the work might have been trying, exciting or dangerous, it was not of the sort that led to miracles.

Cavazzi was, as any reader of his work can easily see, superstitiously pious and easily swayed by the appearance of miracles.  An attestation to this can be found in the chronology of events he placed at the end of MS B.  This chronology is almost entirely concerned with various heavenly signs, comets, unusual births or deaths (often super or preternatural) and the like, along with precise chronological details, although most of it concerns the period prior to 1660.  It appears that before this date, Cavazzi had kept some sort of records of miracles and supernatural events, and perhaps decided after reaching Matamba to change this to a detailed study of this one event, the conversion of Njinga.  Likewise, he is remarkably hazy on chronology of his own movements prior to 1660 (despite his precision on miraculous events), for in MS B he dates his travels to Lubolo with the Portuguese army to 1658, when an examination of other data conducted by Graciano Maria da Leguzzano (editor and translator of Istorica Descrizione) shows that the event took place in 1659.[13] Thus, it appears that Cavazzi’s writing prior to about 1660 was concerned with the small scale daily miracles, and not with either his own daily life or the life of the Africans in whose midst he was working.  His writing (mostly in MS B) about his movement in this period is hazy on chronology, although it does contain vivid detail, suggesting that he was writing from memory about it when he did set it to writing.  In conclusion, we can use the material found in the various MSS to date his writing to the period of 1660-65, for previous periods he seems to have relied on his memory, while during this period the composition of the MSS may have been the only writing he did.  If we can speak of Cavazzi leaving “field notes” of his trip in Africa, there is a very strong probability that these are the text in the Araldi MSS.

From “Missione Evangelica” to Istorica Descrizione

If Cavazzi finished MS A at Matamba in 1665, he was still working on it when he left there, in 1666, for Luanda, where he took up the post of prefect of the Mission on 12 January 1667.[14] We can assume his arrival in Luanda predated 8 September 1666, the date when he seems to have finished MS C in Luanda.  He worked on revisions to the whole text, some of which are dated as late as 1668, when he left Angola for Europe, departing for Brazil on 13 September 1667.[15] Cavazzi returned to Europe, quite sick, in 1668 and only in 1669 had he recovered his health.  It was probably during this time that he approached the Cardinals of Propaganda Fide with the idea of publishing his book on the mission in central Africa, and on 25 March 1669 Cavazzi received an order to produce a full scale history of the Capuchin mission in Africa.[16]

Instead of simply continuing or completing his revisions to the “Missione Evangelica” as one might expect, Cavazzi embarked on a whole new work, which he was calling variously “Descrittione” or “Historia del Congo” in requests that he sent to the archives of the Propaganda Fide for documents.[17] He reported good progress on the work in September 1670, and the whole manuscript was finished before 6 June 1671.[18] This manuscript is no longer extant[19] although we do possess a copy of the contents page, and from it we can see that it is a very close ancestor of Istorica Descrizione as it was published in 1687.[20] Cavazzi fully expected it to be published within a year or so, and in 1673 he began preparations for his return to Africa.  As far as we can tell, he did nothing further to his MS of 1671 after that, although he only died in 1678.[21] He entrusted the manuscript to another Capuchin, Bernardo da Firenze, who continued editing it until his death, when it was finally brought to press by yet another Capuchin, Fortunato Alamandini da Firenze.  The Propaganda Fide decided not to publish the text, due to a shortage of funds (and, no doubt, its imposing length), and its publication history, recounted in detail by Pistoni and Francisco Leite da Faria[22] involved a number bitter personal and family rivalries which delayed its appearance for 16 years.

Both the contents page of the 1671 MS and Istorica Descrizione affirm that Cavazzi’s new history of the Capuchin mission in Africa was organized quite differently from “Missione Evangelica” and rather more like the earlier drafts whose fragments, attached to illustrations, are found in the front matter of Volume A.  The history of the mission, lives of the missionaries, historical accounts of central African kingdoms and descriptions of life and customs are interwoven in Istorica Descrizione, unlike the purely historical and ethnographic account that makes up Volume A in the Araldi collection.  As a result, it is not easy to see precisely where Istorica Descrizione made use of the Araldi text, but once a detailed comparison is made it is clear that virtually all that Cavazzi wrote in the Araldi MSS eventually found its way, often almost verbatim, into Istorica Descrizione.  In addition to new material on Kongo and the Capuchin mission to Benin which Cavazzi obtained either from missionaries or archives in Europe, Istorica Descrizione drew largely from MS B for its section on missionaries, and the text under present consideration, MS A, went largely into Books 1, 2, 4, 6 and 7 of Istorica Descrizione.  Table 1 shows the exact correspondence between the pages of Volume A and their place in Istorica Descrizione.


Pages of Volume A                             Book and paragraph of ID

Book 1

1-57                                                                 II, nos. 1-36

57-69                                                   no counterpart

66-99                                                   II, nos. 37-69 and I, nos.

217-18 (= A, pp. 82-3)

100-101                                                           no counterpart

101-102                                                           I, no. 48

102-103                                                           I, nos. 45, 259

104                                                                  I, no. 319

105-109                                                           II, nos. 70-71

110-125                                                           I, nos. 216-52

125-29                                                 no counterpart

130-139                                                           I, nos. 263-70

Book 2

1-20                                                                 II, nos.

20-6                                                                 V, nos. 106-119

67-73                                                   VI, nos. 1-3

74-8                                                                 V, no. 118

78-107                                                 no equivalent

107-27                                                 VI, nos. 3-22

128-30                                                 IV, nos. 103-109

131-41                                                 VI, nos. 37-60

142-43                                                 no counterpart

143-55                                                 no counterpart, but some

scattered in VI, nos. 93-109

155-66                                                 VI, nos. 95-98

166-84                                                 VI, nos, 63-76

184-224                                                           VI, nos. 101-21

Book 3

1-41                                                                 VII, nos. 35-62

Anyone who compares the passages cited in this table, however, will quickly see that while the general discussion is the same in both works, the exact wording and the information conveyed is somewhat different in each.  Furthermore, neither one or the other is always clearly superior in giving information.  In many cases, the reason for the greater detail supplied in Istorica Descrizione comes from the fact that Cavazzi filled out the earlier work by consulting other sources once he returned to Europe.  This has been abundantly clear in the case of Kongo[23] but it is even the case MS A, area of Cavazzi’s greatest expertise.  Much of the new material found in Istorica Descrizione that is not in MS A is from Antonio Gaeta da Napoli’s book, La maraviglosa conversione della Regina Singa… published in Naples in 1669,[24] which Cavazzi must have seen shortly after his return to Europe.  Since Gaeta was in Matamba before Cavazzi, and was personally responsible for Queen Njinga’s conversion as well as being an eyewitness to events that took place before Cavazzi’s final arrival in Matamba in 1660 (Cavazzi had stayed briefly in Matamba in 1658)[25] it is not surprising that Cavazzi should have favored his accounts for this period over the information he had been able to glean himself.  Cavazzi was much less ready to accept Gaeta’s authority on other matters, however, for he favored his own version of the traditional history of Ndongo over Gaeta’s and relied little on him for ethnographic material as well.

Another group of material found in Istorica Descrizione and apparently lacking in MS A is a fairly substantial account of Cavazzi’s own stay at Njinga’s court found in MS B, where Cavazzi gives his own biography, and naturally enough was apparently in a quandry as to where accounts of Queen Njinga that he was personally involved in belonged.  In “Missione Evangelica” he split his own biography off from the mission in Matamba, while in Istorica Descrizione he joined the two back together.  Still more related to events which took place in 1664 after he finished the main body of Book 2 of MS A, but were incorporated into Istorica Descrizione.

However, even if one subtracts these contributions, there is still a small residual of material found in Istorica Descrizione which differs from and in some cases supplements that supplied in MS A.  An example of this comes from Istorica Descrizione, Book 6, no. 59 which corresponds to a passage in Volume A, p. 141, where Cavazzi and Queen Njinga are watching a festival which involves a great deal of dancing.  In the Araldi manuscript the events are briefly described, while in Istorica Descrizione there is quoted, at length, a conversation between Cavazzi and Njinga on the dancing.  Apparently, in rewriting this section in 1671, Cavazzi decided to adorn it with additional detail, presumably drawing on his memory of the event.  The conversation, recalled nearly a decade after the fact, cannot be as precise as it appears in the published text, and we can only hope that it was not simply an embellishment from Cavazzi’s imagination.  This is only one of many such examples that could be brought from close study of the two texts.

Additional detail comes not only from revisions in Istorica Descrizione, however, for “Missione Evangelica” also contains detail and information which never made it to the printed work.  For example, Cavazzi carefully outlines the kijila (taboos) given to her Imbangala followers by their legendary foundress, Andumba Tembo in the Araldi Manuscripts, under fourteen points, each carefully numbered in both Kimbundu and Italian (Volume A, pp. 25-30).  This section corresponds quite well with Istorica Descrizione, Book 1, nos. 26-30, although the kijila are not individually pointed out and numbered.  The Araldi text goes on, however, to give the observation of the laws three more times, first the way in which they were kept in Kasanje in the 1660’s (Volume A, Book 1, pp. 48-62), then as they were kept by Queen Njinga when she was  “Jaga” (Book 2, pp. 92-107) and then after she became a Christian (Book 2, pp. 143-55), so in this case one gets a better idea of the exact development of these laws from the manuscript than the published version.  In addition, the Araldi manuscripts give full texts of several letters which are only summarized in Istorica Descrizione, especially the letters of various visitors to Kasanje, or the text of various demands put to Kasanje by Christian missionaries (Book 3, pp. 37-40).

On occasion, details in the two texts complement each other to give a fuller story than would be available from either one separately.  A good example comes from the account of a raid by Kasanje on Catole and Dalangue in Matamba in October, 1661.  The Araldi Manuscript (Book 3, pp. 30-1) tells us that it involved 2,000 soldiers and that in the rout of the Kasanje army, Matamba captured 150 prisoners, 16 flags and 16 officers “to go with the flags”, 12 of which held ranks equivalent to captains, and 4 held ranks equivalent to colonel.  None of these details are found in Istorica Descrizione (Book 7, no. 58) where the same raid is described, but the published version does tell us that these two sobas were located near the capital of Matamba and that 200 members of the routed army died by drowning while attempting to cross a nearby river.

There is, therefore, no simple relationship between the two texts, either in organization or in information.  Although from a perspective of total amount of information given neither one can be judged clearly superior, the Araldi text does have several advantages.  First of all, it represents information collected almost entirely by Cavazzi himself, while the published account contains many elements drawn from Gaeta, who might better be consulted first hand.[26] Similarly, many of the additions in the published edition that cannot be attributed to other sources are from Cavazzi’s less than perfect memory, fertilized, perhaps, by his somewhat vivid imagination, while the text of the Araldi Manuscript, written in Africa more or less at the time of the events in question is less likely to be distorted by time.

A special digression must be made to discuss the question of the illustrations, since both the Araldi texts and Istorica Descrizione are illustrated.  In 1987 Ezio Bassani published a careful analysis of all the pictures as well as full color reproductions.[27] The illustrations in the Araldi MSS, 30 in all, are in color and executed in a simple, but direct style that resembles the style of the illustrations used to adorn the Lisbon (Academia das Cienças) manuscript of another great seventeenth century author on Angola, Antonio de Oliveria de Cadornega.[28] There seems to be little doubt that Cavazzi executed them personally, for on the one hand, they are integral parts of the draft manuscript on which they were first drawn, and at times the text in Cavazzi’s hand (as well as legends in the pictures) encroach on the pictures.  Cavazzi has left us an autographed picture in a very similar style elsewhere as well, for he chose to illustrate one of his reports to Rome with a picture showing him and several companions shipwrecked in a small boat.[29]

In addition to the 18 illustrations found in the front matter, other illustrations are found scattered throughout the text, but they are not placed with any particular reference to the material around their location, although those drawn from earlier drafts (where text survives around the illustration) were clearly intended to illustrate the passages in the text.  Since these illustrations are not numbered, Chart Two reproduces the numbering scheme made for them by Ezio Bassani in his color edition of them:


Number Location in Manuscript Subject

Front matter:

1                      Unnumbered first page of front          Jaga Sacrifice


2                      p. 153 of draft (no roman number)     Sixth Trial

3                      IX                                                                                Same sacrifice

(eating flesh raw)

4                      pp. 29-30, or XII-XIII (2 pages)        Njinga’s baptism

5                      p. 37, XIX                                                       Woman having breasts

cut off.

6                      No page, from Book 3, Chapter 1      Smiths working

7                      p. 44, XXXII                                                  Jaga with Lungua

8                      p. 45, XXXIII                                                 Queen leading                                                                                                             retainers

9                      XLII                                                                            Queen, musician? and


10                    XLIV                                                                          Ganga ya Muullaa?

11                    XLVII                                                             Sketch of Njinga?

12                    XLVIII                                                                       Birth of Njinga

13                    no number                                                                   Man with head

chopped off

14                    no number                                                                   Man having chest cut

15                    no number                                                                   Ganga ya Quibanda

16                    p. 42, LVIII                                                    Njinga’s general

looking at crucifix

17                    LIX                                                                             Njinga looking at


18                    LXIII                                                              Ngimbo and Temba


19                    Start of Book 2:                                              First king of Dongo


20                    Book 2, after p. 32                                          A guard sneaking up

on sleepers

21                    Next page                                                                    Singhelli

22                    Next page                                                                    Women with bows and


23                    Next page                                                                    Temba Andumba

pounding baby

24                    Next page                                                                    Njinga becoming a


25                    Next page                                                                    Woman having breasts

cut off

26                    Book 2, after p. 210                            Queen with archers

27                    next 3 pages                                                    Njinga on funeral


28                    next page                                                                     Cruder version of


29                    Start of Book 3                                               Jaga cutting off

person’s head

30                    Book 3, after p. 36                                          Priest speaking

with Lion

As is the case with the texts, there are similarities between these illustrations and the engravings in Istorica Descrizione.  For example, the illustration I have designated number 4 in the Araldi manuscript is very similar to illustration number 43 in Istorica Descrizione (both show the baptism of Queen Njinga).  Clear similarities can also be seen between Istorica Descrizione, number 26 and Araldi numbers 6 and 19.  Whole or partial resemblance can also be found in the following material:

Number in Araldi Number in Istorica Descrizione

1                                                                                             33

2                                                                                             18 (part)

4                                                                                             43

6                                                                                             26

7                                                                                             23 (part)

9                                                                                             44

10                                                                                            34

15                                                                                            31

17                                                                                            45

18                                                                                            29

19                                                                                            26

23                                                                                            30

26                                                                                            44

27                                                                                            47

The Araldi Manuscripts contain 13 illustrations not in Istorica Descrizione while Istorica Descrizione contains 34 illustrations not in the Araldi Manuscripts, most of which pertain to Kongo (which is covered in the Araldi manuscripts only in the unillustrated MS B).

In many ways the painted illustrations, however, are superior in information to those in Istorica Descrizione.  In part this is because some fine detail was inevitably lost in converting a painting to an engraving.  Bassani’s analysis of the paintings and comparisons to material objects found in museums, other illustated collections, or field work in central Africa, shows that Cavazzi had a fine eye for details of designs, color and form.  This is particularly noticibible for cloth designs, which Bassani can document clearly from museum samples of central African cloth collected in the seventeenth century.[30] This particular detail, however, did not come across in the engravings.  The engravings also employ other artistic conventions characteristic of seventeenth century European art (such as the depictation of ancient ruins on hills in the background) which decrease their historical value.

In addition to difficulties arising from the different media (engraving versus painting) the engravings suffer from having been executed by people who had no first hand experience in Africa.  Many were done after Cavazzi left for Africa, for document from the library of the Archgynasim of Bologna indicate that a Capuchin named Paolo da Lorena was entrusted with producing at least some of the engravings–a letter of 31 January 1674 (two years after Cavazzi’s departure) indicate he had finished eight.[31] Eight other engravings are “signed” by Fortunato Alamandini in Istorica Descrizione, the remaining 16 illustrations were probably the work of da Lorena or another unknown engraver.[32]

Given the similarities of Cavazzi’s paintings and the engravings where there are versions in each work, however, it seems reasonable to suppose that the engravings were done from art work that was originally executed by Cavazzi, and unlike some of the illustrations in sixteenth or seventeenth travel literature, were not simply imaginative artist’s impressions of scenes composed from references in the text.  This is more or less confirmed by most of the engravings signed by Alamindini which depict flora and fauna not native to Italy and not described in sufficient detail in the text to produce the accurate pictures.[33] In any case, the fact that the illustrations remain in the Araldi Manuscript suggest that Cavazzi probably executed a whole new set of paintings or sketches for the illustrations (or perhaps used other copies from other drafts, leaving only a few left overs in the manuscript attached as front matter).  Since some of the illustration in the Araldi manuscripts are duplicated (and Cavazzi obviously had no compunction about repeating work) we should not assume that there were once a whole second set of illustrations which the engravers reproduced as best they could.

Thus, the Araldi manuscript allows us to assume the general authorship of all the illustrative material not only in the manuscript, but also in Istorica Descrizione.  These illustrations are less useful for what they purport to show than the details on life and culture they represent.  The picture of Ndongo’s legendary “blacksmith king” Angola Mussuri (no. 6) is an excellent representation of a central African blacksmith at work, with his tools and products surrounding him.  A picture of the birth of Queen Njinga (no. 12) is a striking illustration of the way women gave birth in seventeenth century Africa.  A number of illustrations show Queen Njinga at various times engaged in a variety of activities.  Cavazzi knew the Queen personally, and since he often takes care to show facial details (at least of himself) it is possible that these represent a portrait of her.  Against this however, is the fact that the Njinga of the illustrations is a young woman, while when Cavazzi met her in 1658 she was already 76 years old.  Likewise, the white spots on her hands and face which he mentions in his verbal description of her (Book 2, p. 85) are not visible in the pictures.  But if the pictures of her are not genuine portraits, they are excellent depictations of court life, musical instruments and weapons.

Giovanni Antonio Cavazzi:  Background and Travels

A detailed biographical sketch of Cavazzi is provided by Francisco Leite da Faria and Graziano Maria da Leguzzano in the Portuguese translation of Istorica Descrizione.[34] Likewise, Giuseppe Pistoni has published a useful collection of documents that deal with Cavazzi’s life.[35] In view of this, it is probably unnecessary to go into details of his life.  Perhaps it is enough to say that he was born in Montecuccolo in 1621 of noble parents and entered the Capuchin order in 1632 at the age of 11.  He did not have a particularly good reputation for scholarship and intelligence–first of all he was denied the training to continue on to teaching levels in the order because of his poor scholastic performance and was moreover, almost denied the opportunity for the African mission because some members of his order held a low opinion of his intelligence.[36] He did, however, have a reputation for piety and zeal which eventually prevailed over his detractors.

Modern readers might be inclined to agree with some of these judgments.  His treatment of Classical literature, for example (greatly reduced, by the way, in Istorica Descrizione) is confused and inaccurate, even by the lower levels of Classical scholarship of his day.  Likewise, Cavazzi’s belief in the miraculous intervention of the Divine in daily events was probably considered an anachronism and even a fault by his colleagues, who were moving more towards the more skeptical attitude that would dominate the eighteenth century.  It was this insistence on frequent Divine or Diabolic intervention that proved a telling point in the long delay in publication of his work.  Church authorities took exception to the numerous miracle stories that lace the Araldi Manuscripts and in all probability were just as numerous in the 1671 manuscript which formed the basis for Istorica Descrizione.  The fact that Church authorities ordered the manuscript edited to remove “miracle stories”, and that the published work is substantially less filled with such stories strongly suggests that this was the principal editing work done to the text.  The 1671 manuscript’s table of contents contains reference to Cavazzi’s chronological table, which if it resembled the table in MS B of the Araldi MSS, was full of such miracle stories, was totally absent from Istorica Descrizione.  So too was the long list of miracles (Many very insignificant) that accompanied the building of Njinga’s church (Book 2, pp. 139-41). Curiously enough, the one event concerning the diabolic possession of a snake that detractors of Cavazzi gave as the prime example of his anachronistic attitude towards the supernatural managed to remain in Istorica Descrizione.[37]

Cavazzi’s personality and outlook effected the nature of his research in Africa, and color its presentation.  We have already suggested that his decision to undertake this history may well have been his reaction to the conversion of Queen Njinga, long regarded as the most terrible anti-Christian force in central Africa.  But whatever his reactions to this conversion, Cavazzi was essentially anti-African in his outlook.  Seventeenth century European travellers had a wide range of reactions to central Africa, depending on their own backgrounds and outlooks, as well as the specific situation they found themselves in Africa.  Viewed from the wide spectrum of opinions, Cavazzi was among the most negative.  His views were sufficiently extreme, even in the toned-down language of Istorica Descrizione that its Portuguese-language editor had to apologize to his African readers in 1965, by saying, after a quote from the text on the virtues of Christian Africans, “Such words are not those of a racist”.[38] But the passage concerns Christians of Kongo, a country which greatly impressed Cavazzi (in spite of his initial negative reactions), and the readers of the Araldi text will agree that on the whole he was very reluctant to concede anything to Africans.  His long passage (Book 1, pp. 64-9) on Africans’ inconstancy, his deliberations on whether Njinga would go the Heaven or Hell (Book 2, pp. 201-13), and his general reluctance to allow much even to Christian Africans strongly suggest that he considered Africans inferior, in virtue at least, to many other people in the world.

Cavazzi’s negative views can be strongly highlighted by contrasting them to his contemporary in Njinga’s court, da Gaeta.  Indeed, his onetime companion is among the most positive of the range of views.  While conceding that Njinga surely fell deeply into vice and sin, da Gaeta is more than willing to accept her explanation for this–that the depredations of the Portuguese drove her to these extremes.[39] If Cavazzi sees vice as deeply ingrained in Africans, and perhaps ineradicable; da Gaeta’s view is just the opposite, that Christianity can instantly transform them.  Thus to da Gaeta the miracle of Njinga’s conversion was that worked by conversion to Christianity in a woman who was intelligent, witty, charming and stead-fast, and when converted became an excellent Christian.  Indeed, from the time he met her, da Gaeta describes Njinga in terms almost approaching a saint.[40] Cavazzi clearly believed that she was inherently evil, and for that reason seems to have always had doubts and reservations about the sincerity of her conversion.  Da Gaeta emphasized the humanity of the Africans, Cavazzi regularly describes them as “inhuman”.

These differing views effect virtually all their reporting of events.  Readers of Istorica Descrizione cannot see the difference as easily as readers of the Araldi text because Cavazzi used da Gaeta’s work so heavily in his revisions that some of its tone has carried over.  But the difference is nevertheless there, and shows up in some ways that are otherwise insignificant.  For example, Cavazzi clearly held a very low opinion of African architecture, often comparing houses and even palaces in terms drawn from the barnyard.[41] Da Gaeta, on the other hand, devotes a full page to describing Njinga’s palace in Matamba, which obviously impressed him for its beauty, spaciousness and skill of execution.[42] It is unfortunate that da Gaeta did not spend more time describing material and religious culture of the Mbundu, for it would provide an excellent corrective to Cavazzi’s views on almost everything.  As it is, we must use the material given us on these matters by Cavazzi and make allowances for his views.

For all the difficulties that Cavazzi’s attitudes present, however, he does provide an extremely valuable record of those aspects of African society which interested him.  He was especially interested in history and religion, and his work on both suggests a methodical zeal in collecting information which make him a first rate contributor to our knowledge of seventeenth century central Africa.[43]

Cavazzi was especially interested in history, and made constant efforts to improve his knowledge.  In addition to the traditions of Ndongo and Kasanje recorded in detail in Books 1 and 2, he made a number of attempts to get at the history of other regions as well; for example, when he went to help in the conversion of Ngunza a Mbamba, a ruler in Lubolo in 1658, he managed to record historical information on that area.[44] He also included local stories about the Kasanje capital,[45] and about Mpungo a Ndongo, capital of Njinga’s rival pretenders to the throne of Ndongo.[46] Not surprisingly, he sought historical accounts in the writings of others whose work he used to write Istorica Descrizione, for his section on Kongo history in that book is one of the most detailed on record.[47] Where he recorded the traditions himself, however, he preserved his own account, as his version of Ndongo traditions differs substantially from the version in da Gaeta, in both the Araldi manuscripts (written before he had access to da Gaeta’s work) and in Istorica Descrizione (otherwise revised with da Gaeta in hand).

Unfortunately for modern historians, Cavazzi tried to synthesize his history as well as just record it, with the result that contradictions in his numerous African informants’ accounts are worked out in one smooth narrative.  This can be clearly seen in his attempts to trace the origins of the Imbangala, where he used European travel accounts of other regions to round out “shortcomings” in his informants’ knowledge.  In the Araldi text he relied on João dos Santos’ account of Mozambique and the coincidence between the “Muzimbos” (invaders of that region) and “Zimbo” the founder of the Imbangala to create a wild synthetic tale of migration.  But the story changed in Istorica Descrizione in part, where accounts of Pigafetta intervened to change some of this early history.[48] The process can be followed by a modern researcher with the books in hand, but there is no trace of it in the texts themselves.  He no doubt did the same with the untraceable oral information he received, thus reducing modern historian’s ability to deal with it.

Cavazzi’s observations on religion are perhaps even better than those on history.  Ironically, it was his own type of religiosity and anti-African attitudes which contributed to this usefulness.  Because Cavazzi did not trust African conversions, and believed that their religion was a diabolic manifestation, he became more interested in cosmology–the underlying system of thought behind religious behavior–more than most of his contemporaries and colleagues in central Africa.  In general, Capuchin (and other) missionaries in central Africa were accepting of conversions, and if the convert performed the necessary sacraments and eschewed a rather limited set of practices which were directly contrary to Christian doctrine (such as consultation with local religious actors) they were accepted as Christians.  This practice had developed from the peculiar circumstances surrounding the late fifteenth century conversion of Kongo to Christianity, and Capuchins carried the attitudes and practices of the Kongo mission with them to Angola.[49]

But Cavazzi was not so trusting, and states clearly in Istorica Descrizione (Book 2, no. 58) that his study of Mbundu religious systems was to allow the deceits of false Christians, or the intrusion of local religious belief, to be found out.  Cavazzi’s deeply anti-African attitudes thus led him to deeper study of African religion.[50]

The study of Mbundu religion is not affected too adversely by his own belief that the Devil lay somewhere behind it.  Once one understands that his regular references to the Devil do not refer to an evil force in local cosmology, but rather to Cavazzi’s own explanation for the power of Mbundu religious actors, the remainder of his description is in good accord with studies of modern central African cosmology.[51] Indeed, his own credulity enabled Cavazzi to understand African religious “miracles” in much the same way as he understood his own society’s “miracles”.  Unlike modern observers of African religions, who do not generally believe that African religious actors have any real access to supernatural power, Cavazzi did.  Despite his frequent references to the use of trickery by African priests, “to fill their pockets”, he was sufficiently convinced of some of their power to take it seriously.  For example, when he describes Njinga’s visit to the Jaga Kasa (Book 2, p. 34) with the purpose of murdering her nephew, Cavazzi blithely states that of course the Jaga’s diviners could have known her intent, but the Devil (the source of their power to know this) had his own reasons to hide her intent from them.  Then, caught in the theological problems of his own explanation, he thought it best to add a statement that of course, Blessed God allowed the whole set of events, for without that allowance, nothing can take place.  He might have been able to understand African systems of philosophy, but surely he was no philosopher.

In order to round off a description of Cavazzi’s background, it is appropriate to discuss his travels in Africa.[52] Cavazzi arrived in Luanda on 11 November 1654, although the Portuguese government, suspicious of a Spanish plot on their colony, tied them up in Luanda until year’s end.[53] Like other missionaries, Cavazzi was then sent out to the interior, and most of his travels were centered on one of several bases in the land east of Luanda.  Cavazzi’s first base of operations was the Portuguese fortress of Massangano on the confluence of the Kwanza and Lukala rivers, where he remained from 19 December 1654 until May, 1665.  After that, he was posted to Mpungu a Ndongo, the court of Ngola Hari, the Portuguese-supported contender to the throne of Ndongo, and a sworn enemy of Queen Njinga.  He remained at Mpungu a Ndongo until 1658, during which time he made a number of shorter trips into the surrounding countryside, including travel with the Portuguese army to Kisama in 1655 and 1657 and a shorter trip to Haku just across the Kwanza in March, 1656.  In June, 1658 he was sent on a special mission to baptize the ruler of Ngunza a Mbamba, and returned to Massangano where he remained until 16 October 1658, when he was dispatched to Matamba to replace da Gaeta.  Unfortunately, Cavazzi arrived at Matamba quite sick after an arduous trip, and the Queen, fearing that his death might be a bad omen, returned him to Massangano.  He was sufficiently recovered by June 1659 to accompany the Portuguese army into Lubolo again, and he followed this army across the northern part of the great Angolan central highlands, right up to the border of Bembe, one of the states of the Ovimbundu, only returning to Massangano by way of a round-about route that took him through the eastern Portuguese post of Ambaca and Mpungu a Ndongo at the end of 1659.  In June-October 1660 he made a trip from Massangano to the court of the Imbangala king at Kasanje, but he failed in his principal mission of assuring a conversion of the “Jaga” to Christianity.  Returning from Kasanje, he was immediately dispatched to Matamba, which he reached probably in November 1660.  He remained in the mission of Matamba until January or early February 1664, making two major excursions, one to the islands of Kindongo, Njinga’s former capital in the midst of the Kwanza river (in 1662) and the other to rural areas in Matamba on the eve of the old Queen’s death in December 1663.  After departing from Matamba, Cavazzi was sent on a mission to Kongo’s coastal province of Nsoyo, and seems to have visited some other Kongo provinces before his return to Luanda, sometime in 1665.  Since he states having completed Volume A in Matamba in 1665, he must have gone back there after returning from Kongo and may have remained there until late 1666 when he succeeded Giovanni da Pavia as Prefect of the mission, a post which he took up on 12 January 1667.  He left Luanda, with the completed Araldi manuscripts in hand, on 13 September 1667.


Insert Map One


Angola in Cavazzi’s Day:  Queen Njinga

As we have already suggested, Cavazzi’s main purpose in writing was to explain and describe in detail the miraculous conversion of Queen Njinga.  Book 1, devoted to the history of the Jagas (Imbangala) is intended to show the origin and customs of the people whose rites Njinga adopted, Book 2 is then the historical background of Njinga’s kingdom of Ndongo and, of course, a lengthy biography of her.  Book 3, the history of the Imbangala of Kasanje, is an account of the unsuccessful mission there, and by the side, an example of the horrors of unconverted Jagas.  The story, despite its diversity, therefore centers on the life and times of Queen Njinga.

According to Cavazzi, Njinga was born in 1582, just seven years after a Portuguese expeditionary force under Paulo Dias de Novais reached the Angolan coast.  Although the force had orders to construct forts, control trade and occupy parts of the coast, it was not specifically aimed at conquering the Mbundu kingdom of Ndongo.  Nevertheless, the incompatibility of the goals outlined in Dias de Novais’ charter and Ndongo’s claims to large parts of the coast rapidly led to war, in 1579.  The massacre of Portuguese merchants in Ndongo’s capital in that year is echoed in Cavazzi’s story found in Book 1, pp.     and repeated in Book 2, pp.    , although contemporary documents tell the story quite differently.[54] A fierce war between Portuguese forces, aided by forces from Kongo which had long had designs on Ndongo’s sovereignty and had been claiming authority over the kingdom since at least 1520, led to stalemate after the Portuguese defeat at the battle of the Lukala in 1590.[55]

This stalemate lasted until the first decade of the seventeenth century when the Portuguese came into contact with the Imbangala.  The earliest descriptions of them, by an English sailor named Andrew Battel who was in Portuguese service, date from about 1605 when a Portuguese-Imbangala alliance was just beginning to form.[56] Although there is still debate about the both the remote and the proximate origins of the Imbangala, these early documents (and to some degree Cavazzi’s history of them) show the Imbangala to be resident in the central highlands of Angola, and be formed in tightly organized military bands.[57] From the point of view of the Portuguese they were simply mercenary soldiers who could be paid to fight in Angola and provided a skilled and powerful supplement to the armies of locally raised Angolans and metropolitan Portuguese troops who had been fighting against Ndongo.  Their ritualistic cannibalism and thirst for blood might be disquieting to priests such as Cavazzi, but it did not deter the Portuguese governors from regularly employing them to fight against Ndongo, and their weight, when put on the military scales tipped the balance in favor of Portugal.[58]

The years after 1610 saw numerous Portuguese successes, and Njinga’s father was forced to abandon the traditional capital of Ndongo and take refuge on the Kindonga Islands, and her brother, who came to the throne about 1617 was equally harassed by the formidable forces of his opponents.  But the Imbangala were also opportunistic and like many mercenary armies in Europe or elsewhere, switched sides when the occasion suited them.  The defection of Kaza’s and Kasanje’s Imbangala bands from Portuguese service in these same years helped the rulers of Ndongo to forge their own Imbangala alliance against the Portuguese.[59]

With the death of Njinga’s brother in 1624 a new element was introduced into the situation:  a succession dispute as he died with only a young son as heir.  Njinga probably had an outside claim, on the one hand she was female, and females were generally not allowed on the throne of Ndongo[60] on the other hand she was probably well placed at court and had helped her brother negotiate a peace treaty with the Portuguese in 1622, when she had gone to Luanda, and as a part of the bargain, was baptized as Dona Ana de Souza.  She was closer to the royal blood than other competent rivals, and if we are to believe Cavazzi, one court faction supported her in spite of her sex, and the suspicion that she had murdered her brother in revenge for his own murder of her only son.  She seems to have had little trouble at court, and her chief rival, Ngola Hari ruling at Mpungo a Ndongo, was not only too distant a relation to make a good claim to the throne, but was a Portuguese vassal.  But just as Njinga’s strength at court had overcome whatever opposition there might have been there to her sex, so Ngola Hari’s support from the Portuguese insured him the military power to back his claim.

The civil war between Njinga and Ngola Hari lasted from 1624 until 1629 when Njinga was defeated and driven from the Kindonga islands, while at the same time she was abandoned by the Imbangala, some of whom returned to Portuguese service against her.[61] Finding that she could not fully control Imbangala bands in her service, and her expulsion from her own territory deprived her of access to Mbundu troops, Njinga took the fateful step of becoming an Imbangala herself.  No contemporary source recounts the event, although Cavazzi records it in lurid detail, but it was apparently successful, for she was able to use a new army to conquer Matamba and make it her base from about 1630 onwards.  From this base Njinga continued an intermittent war against her main enemy, Ngola Hari (and with him, of course, the Portuguese) at the same time as she built a major state on the Kwango river.  Sources of the mid-1650’s show her at war in southern Kongo, across the Kwango in the Yaka kingdom, against Kasanje to her south, and against the Portuguese.[62] The high point of her career came during the Dutch occupation of Luanda (1641-48) when she occupied the southern Ndembu region and was able to lay siege to Massangano, the main fortress in Portuguese hands.  It was only the timely arrival of a rescue mission under the general Salvador Correia de Sá in 1648 that saved the day for the Portuguese and forced Njinga to abandon her more ambitious plans.

The years after the restoration of Portuguese power in Angola brought a gradual rapprochement between Njinga and the Portuguese, and from 1654 Njinga began diplomatic overtures to normalize relations with them, obtain a peace treaty and secure the return of her sister, in Portuguese hands since 1646.  Reconversion to Christianity was part of the Portuguese terms, and consequently, Njinga allowed Capuchin clergy, led by da Gaeta, to enter Matamba after 1655.  They remained in Matamba until after Njinga’s eventual death in 1663, when civil war between Christian and Imbangala elements in Njinga’s army forced them out.[63] Given the essentially political and diplomatic nature of Njinga’s granting permission to clergy to enter the country, the apparent fervor and sincerity of Njinga’s embracing Christianity in the last years of her life surprised and pleased the missionaries, and does indeed stand in contrast to the rejection of Christianity by Kasanje, a close ally of the Portuguese since the 1640’s.

It is perhaps inevitable that this convoluted career should lead to romantic interpretations.  Indeed, most of the contemporary accounts stress the marvels of her character and career, comparing her with great female figures of antiquity such as Zenobia, Cleopatra or Semiramis, as we see in Cavazzi;s account, and as can be found in da Gaeta and Cadornega’s as well.  As if a female ruler who was a redoubtable fighter were not enough to make legends, her religious career, from Christian convert of 1622 to ferocious Imbangala cannibal in 1630 to her marvelous return to Christianity in 1656 furthered it.  Other bizarre aspects of her reign, whether it was the unspeakable cruelties listed by both da Gaeta and Cavazzi or her penchant for forcing her male “concubines” to sleep with her serving girls without touching them, only serve to increase the aura of mystery and splendor.

More recent commentary has worked from this background.  Njinga was the subject of a romantic novel published by Coustilhon in 1769 using da Gaeta as a base.[64] Cuvelier and Boon continued the long tradition of presenting her life as an example of the saving grace of Christianity in their Flemish study of 1957.[65] More recently, she has become a nationalist hero, Donald Sweatman turned her story into an edifying tale for African school-children, while Manuel Pedro Pacavira made it an inspiring tale of anti-colonial resistance written while he was himself in prison for his nationalist activities in Angola.[66] She is widely regarded as a proto-nationalist heroine in independent Angola today.

Scholarly study of Njinga has just begun.  David Birmingham and Jan Vansina, pioneers in the study of central African history, have tried to place Njinga’s relations with the Portuguese in a more realistic context, while Walter Rodney and Basil Davidson have seen her as an early nationalist.[67] Only Joseph C. Miller, whose brief study of Njinga appeared in 1975[68] has tried to place her reign in the context of African politics as well as Euro-African relations.  None of these studies have made use of the full range of source material available for the study of her life, however, as Beatrix Heintze’s detailed study of Angolan history in the period 1617-1630 (with detailed evaluation of her earlier history) based on voluminous but previously unused documentation left by the Portuguese governor Fernao de Sousa and now found in the Biblioteca da Ajuda in Lisbon shows.[69] A fuller biography of her must make use of da Gaeta’s volume as well, a rare book which is of fundamental importance and has not been used by any of the above scholars.  It goes without saying that close study of the present text is necessary as well.  Njinga’s life is unusually well documented, including a large corpus of letters written by her (or on her behalf, as her knowledge of writing is uncertain) and could probably be discussed at length.

Presentation of the Text

I have produced a typed Italian transcription of the entire text of volume A, taken from a microfilm copy of the text made in 1976 by Joseph C. Miller.  The original text contains only one lacuna of one page (pp. 20-21) and for the most part is written in a clear hand that is quite legible.

After examining the front matter, I decided not to transcribe the material, as it would add little to the text as now established.  Although the exact language is sometimes quite different from equivalent passages in the main text, a laborious comparison of the two would not substantially change the content.

The text includes a number of crossed out words and phrases, as well as occasional marginal notes.  I have included these in the translation, placing the crossed out material in square brackets immediately following the text as Cavazzi wished it to read after emendation.  Where marginal notes are clearly linked to the text (by an asterisk like convention) I have added it in the text, where there is not a clear indication, I have placed it where it seems to belong.  This material is also in square brackets introduced by the words “marginal note” or “in margin”.


Unlike Cavazzi’s writing hand, his language was most difficult.  The problem is two fold.  On the one hand, Cavazzi almost completely eschewed meaningful punctuation, a feature common of seventeenth century texts, although extreme in this one.  Moreover, he often left pronouns with unclear antecedents in these overlong sentence, so that translation becomes interpretation as well.  On the other hand, Cavazzi’s writing style seems to have been deliberately flowery, especially in that he attempted at times to make Italian into Latin.  An extreme example of this comes from the legend of illustration no. 20, which says “Voluntarie o involuntarie, con pena da morte quello punisce” (Literally, “voluntarily or involuntarily, with pain of death this one is punished”)  These problems require an unusually high degree of attention to fine points of grammar, and even then produces sentences that are all but incomprehensible.

Another difficulty in Cavazzi’s text is his own somewhat muddled sense of structure.  He goes off to say something, gets caught up in long asides, and forgets his original intent which is left forever hanging.  His attempts to be rhetorical, or to introduce Classical and Biblical allusions also interfere with a clear interpretation of the text, since many of his allusions are impossibly obscure.  Cavazzi’s original editor complained of most of these faults in his language in the introduction to the first edition, and as we can see his solution to many of them was simply to eliminate the difficult passages altogether.  In spite of his efforts, however, Istorica Descrizione suffers from some of the same faults of the manuscript.  It was for this reason that Garziano Maria da Leguzzano, in making the Portuguese translation of 1965 decided to make a translation that was really an extended paraphrase, as a reader who consults text and translation rapidly discovers.[70]

Our translation, as any translation of this text, must therefore be used with caution.  We have made no attempt to imitate Cavazzi’s style and have introduced punctuation freely, even when uncertain that Cavazzi intended for the phrases to be so divided.  On the other hand, we have tried to translate and not to paraphrase, even if at times, it makes the text a bit hard to follow.  The usual proviso of translators, that the original ought to be consulted at every point by any serious reader, goes full measure in the case of this translation.

The translation has a somewhat complicated history.  Thornton produced a provisionary translation of Book 1 and the first forty pages of Book 2 in 1982-3, but fearing that the translation was inadequate, obtained help from Maria Luisa Martini who translated Book 2, pp. 40-100 in 1983.  In 1986, Beckingham joined the project, and produced a translation of the remaining portions of Book 2 and all of Book 3, then subsequently emended and in many places completely retranslated the earlier work of Thornton and Martini.

Orthography and Conventions

The text employs the original orthography of Cavazzi’s Italian text for all proper names and untranslated foreign words, original language equivalents being supplied in annotations where appropriate.  In the apparatus, when dealing with European languages, I have tried to employ spelling of terms appropriate to the modern form of the language–thus the Portuguese name João is so spelled, instead of various seventeenth century possibilities such as Joham, and I write Salvador de Sá instead of Salvador de Saa.

Kimbundu words, however, require further explanation.  Texts have been written in Kimbundu since 1642, when the Jesuit missionary António de Couto edited the Kimbundu catechism composed by Francesco Paccoino, who had labored on it from 1626 until his death in 1641.[71] Cavazzi, however, had his own orthography for words in Kimbundu, influenced by the usage of the catechism, which he had probably seen, but also influenced by Italian.  For example, he used /qu/ (which would normally represent a /kw/ sound in Italian) to equal the /k/ sound in Kimbundu as in the Portuguese orthography of the catechism.  Thus, he writes “Cassanji Caquingurij”, intending it to be pronounced “Kasanji Kakinguri”.  On the other hand, his spelling of the Kimbundu term xingila (priest) as “Singhilla” shows the Italian orthography of the /g/ sound, rather than the Portuguese.

Until 1980, Kimbundu was written in a variety of orthographies, most of which, like Paccoino’s were based on Portuguese phonology, although some, especially those produced since the late nineteenth century by Protestant missionaries from Switzerland or the United States, sought to adopt orthographies based on international systems of “one letter, one sound”.[72] In 1980, the Angolan Instituto Nacional das Linguas produced a carefully reasoned orthography for Kimbundu, which the government of Angola intended to make standard.[73] For the most part, I have sought to use this orthography, although without the tone markings and reduplicated long vowels that the Instituto recommends.[74]

Unfortunately, the Instituto orthography is not entirely adequate.  For one thing, tone is not fixed (hence my decision not to mark it) in sentences or phrases and may shift as words are ellided.  Secondly, the Instituto’s orthography provides no rules on how to handle the frequent ellisions in Kimbundu, since they focused their work on words rather than sentences or phrases, and affect both tone and vowel length (hence my decision to eschew marking vowel length as well).  The variety of ways in which this feature of Kimbundu is manifested can be seen in texts produced in the language, or even the standard orthographies of Kikongo, a closely related language with several different national orthographies (Congo-Brazzaville, Zaire and Angola).

I have decided to write Kimbundu using ellision rules similar to those from the Zairean orthography of Kikongo, which generally does not display ellisions in the text, but divides words according to grammatical use rather than pronunciation.  This differs from the seventeenth century conventions:  in catechisms, for example, the Kikongo catechism of 1624, and the sermon that accompanies the dictionary of 1650, the writers chose to restructure word division to reflect tone bridging and ellision, both common features of west central African languages.[75]

The Kimbundu catechism of 1642, on the other hand, makes more liberal use of the apostrophe to connect words and reflects the complicated feature of vowel changing that accompanies ellision.  Cavazzi wrote few phrases, and rarely shows how he solved the problem, but he seems to have adopted the solution of the Kikongo texts, especially in writing names:  like many of his contemporaries, Cavazzi routinely joined conjunctive particles with the second word in a phrase, which reflects their pronounciation rather than grammatical relationships.  Thus, the name that is often written Kasanje ka Kinguri in modern texts (reflecting the Zairean solution) is spelled by Cavazzi “Cassangi Caquingurij” reflecting quite accurately the pronunciation of the phrase.

I have decided on this orthography as a result of existing usage in historical literature, especially that in English, as much as a conviction that it is the best solution to the orthographic problems.  Some influential work on the history of Kimbundu speaking areas has been done by writers, such as Jean Cuvelier, who spoke Kikongo rather than Kimbundu, and so adopted the Kikongo orthographies.  It was probabaly Cuvelier who popularized the Kikongo spelling (and pronunciation) of Queen Njinga’s name as “Queen Nzinga” which then became widespread in literature in both French and English about Ndongo’s famous queen.  Portuguese language material, however, has retained the Kimbundu orthography, writing either “Njinga” or “Jinga”.[76] My orthography accepts Cuvelier’s approach to the orthography of phrases, but adopts the Kimbundu alphabet of the Instituto.


[1].  Giuseppe Pistone, “I manoscritti `Araldi’ di Padre Giovanni Antonio Cavazzi da Montecuccolo,”  Atti e memorie, Accademia Nazionale di Scienze, Lettere e Arti di Modena 9 (1969):  152-65.

[2].  See editor’s introduction to Istorica Descrizione de tre regni Congo Matamba ed Angola (Bologna, 1687).  There are numerous translations and subsequent editions, see modern critical Portuguese translation of Graziano Maria da Leguzzano, Descricao Historica dos tres reinos Congo, Matamba e Angola (2 vols., Lisbon, 1965).

[3].  Teobaldo Filesi and Isidoro de Villapadierna, La `Misio Antiqua’ dei Cappuccini nel Congo (1645-1835) (Rome, 1978), pp. 114 and 150-3.

[4].  Pistoni, “Manoscritti `Araldi’”, p. 159.

[5]Ibid, pp. 158-61.   A facsimile copy of the illustration of Cavazzi’s shipwreck, from the Archives of the Propaganda Fide in Rome is found in Francisco Leite da Faria’s introduction to da Leguzzano, Descricao Historica, p. L.

[6].  John Thornton, “New Light on Cavazzi’s Seventeenth-Century Description of Kongo,” History in Africa 6 (1979): 253-4.

[7]Istorica Descrizione, Book    no.       citations to book and paragraph number allow readers to consult the original 1687 edition (available on microfiche from Interdoc, Zug, Switzerland) or the Portuguese translation, da Leguzzano, Descrição Histórica.

[8].  Thornton, “New Light”, pp. 258-60.

[9].    Most important in Volume B is the detailed chronicle of events and Cavazzi’s own biography (although the latter is for the most part found in Istorica Descrizione).

[10].  See Book 6, nos.

[11].  This book was published posthumously, in “a historical style by Francesco Maria Gioia da Napoli” from a “Relatione sent to him by Antonio da Gaeta” as La maravigliosa conversione alla Santa Fede di Cristo della Regina Singa (Naples, 1669).  Antonio da Gaeta had apparently finished writing the account by June, 1658, as he refers to his “Relatione” in a letter to the Secretary of the Propaganda Fide of that date, António Brásio, (ed.) Monumenta Missionaria Africana (13 volumes, 1952-82) 12: 160-63.  Bibliographers have occasionally mistaken Gioia da Napoli as the author rather than the editor of the text.

[12].  See letter cited in note 11, and da Leguzzano, Descrição, 2:  381-2.

[13]Istorica Descrizione, Book 6, no. 43 (and footnote in da Leguzzano Descrição).  The same date is in MS B, p. 517.

[14].  da Leguzzano, Descrição, 2: 432.

[15]Ibid, loc cit.

[16].  Reproduced in Istorica Descrizione, Book 7, no. 164.

[17].  Documents reproduced in Brásio, Monumenta 13: 106-7 (“Descrittione”) and 108 (“Historia del Congo”)

[18].  Cavazzi announced its completion in a letter to the Cardinal Prefect, 6 June 1671, Brásio, Monumenta 13: 133-4.

[19].  Thornton, “New Light”, pp.? 254 and 261, note 8.

[20].  Giuseppe Pistoni, Fra Giovanni Antonio Cavazzi da Montecuccolo:  Documenti inediti (Modena, 1972), pp. 5-9.

[21].  Leguzzano, Descrição 2:  432; Thornton, “New Light”, p. 254.

[22].  Pistoni, Cavazzi; Francisco Leite da Faria, introduction in da Leguzzano Descrição, pp. XXXVI-IX.

[23].  Thornton, “New Light”, pp. 257-60.

[24].  See note 11, above.

[25].  da Leguzzano, Descricao, 2: 431.

[26].  Da Gaeta’s book is, however rare.  The only copy found in the United States is at Northwestern University and is too fragile to film.  I have consulted the copy at the Bibliothèque Nationale, Paris.

[27].  Ezio Bassani, “Un cappuccino nell’Africa nera del seicento.  I desegni dei Manoscritti Araldi del Padre Giovanni Antonio Cavazzi da Montecuccolo,” Quaderni Poro 4 (1987), with a very detailed annotation on the material culture of the Mbundu displayed in the pictures.

[28].  See the black and white reproductions of Cadornega’s Lisbon manuscript in the modern critical edition of José Matias Delgado and Manuel Alves da Cunha, História geral das guerras angolanas (3 vols., Lisbon, 1940-42, reprinted, 1972).  A few of these pictures have been published in full color in Comissão Nacional para as Comemorações dos Descobrimentos Portugueses, Portugal na abertura do mundo (2d ed., Lisbon, 1990), pp. 47, 62-3, 66-7, although the reproductions are unfortuately quite small.

[29].  da Leguzzano, Descrição, p. L.

[30].  Bassani, “Cappuccino nell’Africa,” pp. 41-56.

[31].  Leguzzano, Descrição, p. XXV.

[32]Ibid, p. XL.

[33].  Cavazzi’s section of flora and fauna is found in MS B of the Araldi texts.  Some versions of the 1687 edition of Istorica Descrizione also contain an unnumbered panoramic view of the rock of Pungo a Ndongo which could only be the work of an eyewitness, and is one of the earliest landscapes of central Africa.  This plate is found in the copy in the New York Public Library, Rare Books section, but was not reproduced in Leguzzano’s edition, nor is found on the Interdoc microfiche edition.

[34].  Leite da Faria’s sketch is in the introduction, da Leguzzano’s is in volume 2, pp. 330-2? (an index of biographies).

[35].  Pistoni, Cavazzi.

[36].  da Leguzzano, Descricao, pp.

[37]Ibid, pp. XXXVI-IX.

[38]Ibid, p. IX.

[39].  Da Gaeta, Maravigliosa conversione, pp. 95-106.

[40]Ibid, pp. 104-06; 264-66.

[41]Istorica Descrizione, Book 6, nos.    and Book 1, no.

[42].  Da Gaeta, Maravigliosa conversione, p. 98.

[43].  Cavazzi had an interest in botanical and zoological novelties of central Africa as evidenced by his last two chapters in Volume B, printed more of? less the same in Istorica Descrizione, Book 1, nos. 53-153.

[44]Istorica Descrizione, Book  , no.

[45]Ibid, Book 7, no. 35.

[46]Ibid, Book  , no.

[47].  Thornton, “New Light”, pp. 254-5.

[48].  See notes to this text, Book 2, pp. 4-7.

[49].  John Thornton, “The Development of an African Catholic Church in the Kingdom of Kongo, 1491-1750”, Journal of African History 25 (1984):  147-67.

[50].  Whatever his underlying attitudes, however, Cavazzi’s own practices in Matamba went a long way to accommodate local religious belief and differed little from that of his colleagues:  This text, pp. 159-66, most of which is not found in the corresponding passages of Istorica Descrizione, Book 6, nos. 95-6.

[51].  See notes to Book 1, pp. 70-101.

[52].  I have followed the reconstruction of da Leguzzano, Descricao 2: 430-2.

[53].  This episode is dealt with in detail in da Gaeta, Maravigliosa Conversione, pp. 49-66, who arrived at the same time.

[54].  The standard treatment of these years is David Birmingham, Trade and Conquest in Angola (London, 1966), pp.

[55]Ibid, p.

[56].  Obtained from him, in manuscript, by Samuel Purchas and edited in Purchas His Pilgrimmes (London, 1625), this text and earlier writing made by Purchas from interviews with Battel are gathered together in a critical edition, old but still useful, of E. G. Ravenstein, The Strange Adventure of Andrew Battel in Angola and Adjoining Regions (London, 1901, reprinted 196 ).

[57].  See Joseph C. Miller, Kings and Kinsmen:  The Imbangala Impact on the Mbundu of Angola (London, 1976).  For a critique, C. C. Wrigley, “Myths of the Savanna”, Journal of African History 15 (1974): 131-5; and John Thornton, “The Chronology and Causes of Lunda Expansion to the West, c. 1700-1852”, Zambia Journal of History 1 (1981): 1-6.

[58].  The most detailed study of this period is Beatrix Heintze, “Das Ende des Unabhangigen Staates Ndongo (Angola).  Neue Chronologie und Reinterpretation (1617-1630)” Paideuma 27 (1981):  197-273.

[59].  Heintze, “Ende”, pp. 265-66.

[60]Ibid, pp. 219, 222-24.  For further details and analysis of this period and Njinga’s claims, see John Thornton, “Legitimacy and Political Power in Central Africa:  Queen Njinga, 1624-63,” Journal of African History 30 (1991)

[61]Ibid, p. 266.

[62].  Birmingham, Trade and Conflict, pp.     see also Joseph C. Miller, “Nzinga of Matamba in a New Perspective,” Journal of African History 21 (1975):  201-16.

[63].  Miller, “Nzinga” pp. 214-16.

[64]Zingha, reine d’Angola.  Histoire africaine en deux partes (Bouillon, 1769).

[65]Koningin Nzinga van Matamba (Brugge-Bussum, 1957).  See also the scathing review of this book by Jean Stengers,

[66].  Donald Sweatman, Nzinga, A Queen Who Saved Her People (London, 19  ) Manuel Pedro Pacavira, Nzinga Mbandi Romance (Luanda, 1975 and second edition undated, appeared before 1979).

[67].  Jan Vansina, Kingdoms of the Savanna (Madison, 1966), pp. 129-30, 134-8 and 142-6; Birmingham, Trade and Conquest, pp. 78-132; Walter Rodney, “European Activity and African Reaction in Angola”, in T. O. Ranger (ed.) Aspects of Central African History (London, 1968), pp. 56-9; Basil Davidson, Angola’s People:  In the Eye of the Storm (New York, 1973), pp. 86, 91-2.

[68].  Miller, “Nzinga”.

[69].  Heintze, “Ende”.

[70]. da Leguzzano, Descricao, p.     see the critique of his translation in Beatrix Heintze,

[71].  António de Couto (ed.), Francesco Paccoino, Gentio de Angola sufficientimente instruido nos mysterios de Sagrada Fede… (Lisbon, 1642).

[72].  The first dictionary of Kimbundu was published in 1805, by Bernardo da Cannecatin, and put together in around 1795, a revised edition was published in 1845;  modern dictionaries and treatises begin with Héli Chatelain

[73].  Instituto Nacional das Linguas, Histórico sobre a criação dos alfabetos em linguas nacionais (Luanda, 1980).

[74].  The Instituto’s decision to insist on tone and vowel length marking is defended with good reason by reference to the minimal pairs in the language, in which the only difference between two words is tone or vowel length.  It must be said, however, that such minimal pairs are fairly rare in Kimbundu, and can usually be unambigously resolved by reference to context.

[75].  The catechism of 1624 is reproduced in a modern edition with a French translation and re-written Kikongo text in François Bontinck and D. Ndembe Nsasi, Le catéchisme kikongo de 1624.  Réédition critique (Brussels, 1978); the sermon is found in Biblioteca Nazionale Centrale Vittorio Emmanuele II, Rome, Fundo Minori 1896, MS Varia 274, “Vocabularium congoese…”  It was published in a re-written version in Joseph van Wing and J Penders, Le plus ancien dictionnaire bantu (Louvain, 1928).  In both editions, the task of re-writing the Kikongo text (for clarity for modern readers) is not only to replace Portuguese orthography (ie /k/ for /qu/) but also to break up the long words and irregular joining of words created by this seventeenth century convention.

[76].  These two spellings highlight another problem in Kimbundu orthography:  what to do about initial nasals before /s/, /z/, /j/, /b/, /p/ and /t/.  In normal speech, such nasals are routinely not pronounced with the word is in phrase initital position, while it is always pronounced when following an article or ellided with another word.  Thus, the queen’s name might be pronounced /jinga/ in phrase initial position, and /njinga/ at other times.  Convention of the Zairean orthograpghy of Kikongo dictates that such nasals always are written, even if not pronounced, which I will follow, again as a concession to widespread usage.