Phoenicians in Brazil?

036

037

Of the recurring, often bizarre attempts to find ancient Semitic inscriptions in the western hemisphere, the most prominent and frequently cited concerns the so-called Paraiba inscription from Brazil.

The Paraiba inscription is said to be a Phoenician inscription carved in stone and found in northeastern Brazil in 1872.

After raising a minor stir in the 1870’s, the Paraiba inscription was almost forgotten—until another copy of the inscription was publicized in the 1960’s by Professor Cyrus Gordon, then of Brandeis University and now at New York University. Based on this newly found copy, Gordon, in 1968, published an article declaring the inscription to be authentic. In a three-column headline, the New York Times declared “Phoenicia Linked to America.”

The contents of the inscription are remarkable—although a skeptic might suggest that it is precisely what we should expect from a Phoenician inscription found in Brazil!

The text recites the story of ten Phoenician ships manned by Sidonians that sailed from Ezion-Geber, down the Red Sea, around the southern tip of Africa, and then headed north, presumably with the intention of sailing through the strait of Gibraltar, into the Mediterranean Sea and then to Phoenicia on the eastern coast of the Mediterranean. Off the coast of Africa, however, one of the ships was separated by a storm and was eventually cast “on this distant shore.” On board were twelve men and three women.

According to Professor Gordon, the Paraiba inscription contains linguistic features which were unattested in Phoenician until after the Paraiba inscription was published. Unless the purported forger could have anticipated these recent discoveries, the inscription must be authentic, Gordon argues.

The Paraiba inscription was first brought to public attention in a preliminary announcement in 1874 by Ladislau Neto. Dr. Neto, who was director of the National Museum in Rio de Janeiro, had been given the inscription by Candido Viana, the head of Brazil’s most prestigious academy, The Brazilian Institute for History and Geography. Viana commissioned Neto to study and publish the inscription. Neto’s preliminary announcement in 1874 was intended to alert the scholarly world to the inscription’s existence.

Neto promptly set about investigating the source of the inscription. Viana had received the inscription in a letter which stated that the original inscription, carved on a stone, had been discovered in 1872 by a plantation slave in Pouso Alto, a village near the Paraiba River in northeastern Brazil. The copy of the inscription contained in the letter was said to have been made by the son of the plantation owner, who was apparently a skillful draftsman. The letter to Viana was signed Antonio Alves da Costa.

Despite tireless efforts, Neto was never able to locate or even to identify Senor da Costa. In short, the trail ends with da Costa. No one has ever seen the original Paraiba inscription, or even its find spot.

When every effort to identify da Costa failed, Neto repudiated the inscription as a hoax and never gave it formal publication.

However, Konstantin Schlottmann published a facsimile of Neto’s copy in 1874, and, although he argued for its authenticity, the inscription was then more or less forgotten. Mark Lidzbarski, the first fully competent epigraphist to deal with the inscription, dismissed it in his Handbook of Northwest Semitic Epigraphy (1898) as an obvious forgery.

The long-discredited hoax was given a new life, however, at a benefit sale for a women’s college. Dr. Jules Piccus of the University of Massachusetts at Amherst was browsing through odd items, and 038noticed an interesting-looking scrapbook which on examination turned out to have belonged to Wilburforce Eames who, at the end of the last century, was head of the New York Public Library. Piccus bought the scrapbook for a dime or a quarter. He can’t remember which.

When he got the scrapbook home, Piccus discovered in it a letter from Neto to Eames which contained a transcription Neto had made of the Paraiba inscription.

Piccus sent a Xeroxed copy to Gordon in November, 1967. Gordon proceeded to declare the almost-forgotten inscription genuine.

Since that time, it has been a principal crutch of those who argue that the ancient Semites sailed across the ocean and reached the western hemisphere long before Columbus. Indeed, Gordon goes so far as to suggest that “the excesses in human sacrifice in Pre-Columbian America (that went to unbelievable extremes among the Aztecs) may reflect the zeal of converts to Phoenicianism,” who presumably brought the practice of human sacrifice to western shores in the voyage described in the Paraiba inscription.

Our understanding of Phoenician and related Semitic languages has increased enormously in the past 100 years. It would be surprising if a modern scholar in control of these materials were unable to detect a Phoenician forgery of a hundred years ago. In fact, recent scholarship provides abundant evidence that the Paraiba inscription is a forgery.

Naturally much of this evidence is technical and requires an understanding of several specialized disciplines.

But it is possible to give the lay reader an understanding of these scholarly tools and the way they are used to detect a forgery. We can also share with the reader some examples of the kind of evidence which establishes the forgery of the Paraiba inscription.

We have already mentioned the fact that no scholar has ever seen the Paraiba inscription, and its absence has never been satisfactorily explained. Nor has the man who brought the alleged inscription to the scholarly world—Senor da Costa—ever been identified. Another suspicious circumstance is the fact that every letter is clear in Neto’s copy. We must suppose—if the inscription is authentic—that the original is in a perfect state of preservation and that the draftsman who copied it was either knowledgeable in Phoenician or had magical powers. In the real world of Semitic inscriptions this virtually never happens—especially in an inscription of this length. These facts, however, only make us suspicious; they do not prove the case. We shall soon turn to some hard evidence. But first, some background.

The scholar’s aim is to understand an ancient civilization. At the outset, however, he must separate 039his materials into chronological periods, lest he use material applicable to one period in order to understand another. (See Paul Lapp’s excellent article on “The Importance of Dating,” BAR 03:01.) That is why dating is so important a part of the scholar’s endeavor.

Fortunately, almost all cultural artifacts develop and change over time. This is as true of ancient inscriptions as it is of medieval armor. As scholars learn more and more about a particular kind of cultural artifact, they can arrange the examples in a chronological sequence based on certain changes in tell-tale characteristics, thereby dating the artifacts relatively, or relative to one another. Eventually, the sequence is dated absolutely; that is; to a particular time period rather than solely in relation to other artifacts in the chronological sequence. These sequences are the basis for typologies which are used for dating purposes.

In Near Eastern archaeology, the most important typology is ceramic typology—pottery. With the development of ceramic typologies—first relative typologies and then absolute typologies—archaeologists have been able to date excavation levels on the basis of ever-present and indestructible potsherds. From a so-called diagnostic sherd—a rim, base or handle—a modern archaeologist is often able to reconstruct the entire vessel and, from this, to date it.

041

Similarly, modern scholars have learned to date ancient Semitic inscriptions. As the ceramicist looks at the nature of the ware, the decoration, the shape, the lip on the rim, or the height of the base, the expert in Semitic inscriptions looks at the shape of the letter (paleography), the spelling of the words (orthography) and the structure of the language (linguistics).

Each of these disciplines is a highly developed specialty with its own typologies capable of dating an ancient inscription. Moreover, each of these disciplines, especially paleography and orthography, was at a very rudimentary stage in the 1870’s when the Paraiba inscription was supposed to have been discovered.

As a result, the forger of the Paraiba inscription—using 19th century scholarship—made terrible blunders in each of these areas, mixing characteristics from different periods in an impossible fashion.

Semitic paleography has been placed on a sound scientific basis only in the last quarter of a century. The shape of the letters (each of the letters develops separately), their stance (the slant of the letters), the kind of pen that was used and the way the pen was held, whether the letters were written above or below an imaginary (or real) line, how thick the strokes are (shading)—each of these tell-tale characteristics change with the passage of time and enable today’s paleographer to date an inscription.

The failure of the Paraiba forger to utilize these characteristics in a consistent manner immediately reveals his fraudulent credentials.

Some letters in the Paraiba inscription (alef) would indicate that the inscription dated from the ninth century B.C., other letters (yod) from the eighth or seventh centuries B.C., still others (he, tsade, and qof) from somewhere between the eighth and sixth centuries B.C., and others (bet, dalet, kaf, mem, samek, taw) from the fourth century or later. The form of the lamed can be found no earlier than the first century B.C.

The Paraiba inscription thus presents us with an impossible mixture of letter forms from various periods, covering a span of time 800 years long.

It is not difficult to detect the source of the forger’s errors.

The basic decipherment of Phoenician was the accomplishment of the great Hebrew grammarian Wilhelm Gesenius (1786–1842). His book on Phoenician inscriptions includes a so-called script chart, that is, a chart giving examples of the forms of Phoenician letters from the inscriptions Gesenius studied. In the nineteenth century, two other scholars, Auguste Celestin Judas and Francois Bourgade, brought Gesenius’ work forward and made script charts of their own, based principally on Gesenius.

Although grounded in the best scholarship of their day, these nineteenth century script charts are useless to a paleographer today. The nineteenth century script charts are based on hand copies of the original inscriptions; accordingly, they are less accurate than modern copies of these same inscriptions. (Modern copies of inscriptions are traced from photographs and even the tracings are corrected on the basis of observed paleographic details in the original.) Scarcely a letter in these nineteenth century script charts would be considered accurate by today’s standards. One common characteristic, apparently repeated by Judas and Bourgade from their use of Gesenius, is that almost all the letters are squatter than they are in the originals.

Another reason these script charts are paleographically useless to the modern scholar is that the charts were not made to disclose diagnostic characteristics. Thus, the examples of each letter were arranged vertically so that it is impossible to observe the height of one letter relative to another. Similarly, no attention was paid to stance or shading. Finally, no attempt was made to arrange the letters in the script charts in any chronological order. Phoenician letters were simply Phoenician letters.

Many of the letters in the Paraiba inscription have paleographic features which cannot be found in any Phoenician inscription; the closest parallels are to the 19th century script charts—where the forger obviously got them.

Frequently, letters in the Paraiba inscription are in the wrong stance. The same error is found in the script charts, whose makers, like the forger of the Paraiba inscription, saw no significance to letter stance. Thus the Paraiba qof leans strongly to the right. There is no parallel to this in Phoenician inscriptions. It is, however, the stance of the qof in the script charts of Gesenius and Judas.

Professor Gordon, in his analysis of the inscription, tells us only that “The script is Sidonian, close to the Eshmunazar inscription (N.B. the shin) of the early fifth century B.C. However, the zayin and yod are more archaic and suggest a sixth-century date.” This is the sum total of his paleographic analysis. In response to my detailed paleographical analogies, Gordon has 042stated only that different scholars vary in their dating and that I present an excessively precise picture. It is true that judgments among competent scholars may vary somewhat, but nothing like the eight centuries represented in the letters in the Paraiba inscription.

Orthographic features (spelling) of the Paraiba inscription also betray the forger.

For the most part, Semitic alphabets consist solely of consonants. Sometimes, however, certain consonants are used as vowel markers. When so used, these consonants (he, waw, and yod) are called matres lectionis (singular: mater lectionis). This Latin phrase, meaning “mother of reading,” accurately suggests that these usages were the first rudimentary efforts at vowel signs.

Conscious use of matres lectionis appears very late in Phoenician—not until the third century B.C., and then sporadically. But the Paraiba inscription, which Gordon dates to the sixth century B.C., is full of matres lectionis.

How did the forger happen to make this error? The reason is simple. In the nineteenth century, the preponderance of Phoenician inscriptions were from the late Punic period (third century B.C. and later) at which time the use of matres lectionis was common. In addition, when the Paraiba forger wanted to use a word in his “Phoenician” inscription, he went to Hebrew inscriptions. Matres lectionis appear in Hebrew much earlier than in Phoenician—as early as the ninth century B.C. when, for example, –y, –w, and –h are used to mark final long vowels: –i, –u, –a. We find some spelling in the Paraiba inscription that might be characteristic of a sixth century Hebrew inscription but not, as we now know, a sixth century Phoenician inscription.

This is not the only orthographic error by which the forger trapped himself. In general, the orthography of the Paraiba inscription is an impossible potpourri. On one extreme, it is classic Phoenician (eleventh to fourth centuries B.C.); on the other, it is neo-Punic (after 146 B.C.). Mixed in are Hebrew spellings which do not occur before the fourth century B.C.

Thus from the orthographic viewpoint, the Paraiba inscription represents no one period either in Phoenician or in Hebrew. It is impossible to place in the tenth century, the sixth century or in the Roman era. The only explanation of this combination of styles and periods is that it was written in the nineteenth century of our own era.

Finally, today we know more about the Phoenician language itself than we did 100 years ago. We know much the Paraiba forger did not know.

For example, in ancient Semitic languages, there is something called a consecutive or conversive verb form. When a verb is used with the particle, the tense of the verb changes. This verb form disappeared in Phoenicia by the tenth century B.C. The Paraiba forger knows this verb form from the Hebrew, where it persisted much longer. But finding conversive verbs in a sixth-century Phoenician text is like finding a Chaucerian word in a twentieth century newspaper.

These are examples of the myriad ways that the Paraiba forger has unwittingly revealed himself. Perhaps it would be possible to question or explain one or two of these tell-tale signs. It is true that all of the disciplines involved—paleography, orthography, and linguistics—involve delicate judgments. Applying these principles is at times as much an art as a science. But the cumulative effect of the Paraiba forger’s blunders is devastating. Under the cold light of modern scholarship, the Paraiba inscription is a plain fraud.

But what of Gordon’s claim that the Paraiba inscription anticipates advances in our knowledge of Phoenician which were unattested when the Paraiba inscription was first published?

This is simply not true. Everything in the inscription was available to the forger in nineteenth century handbooks or from uninspired guesses based on these easily available sources.

043

For example, Gordon tells us that the singular mt is used for “man” in the Paraiba inscription and this form was not known until 1933 when a Ugaritic text with mt meaning “man” in Canaanite was first published. Although mt meaning “man” was first found in 1933 in an Ugaritic inscription, mt (“handmaid”, derived from mt, but misunderstood) was well known in Punic. Nineteenth century scholars misunderstood the form and incorrectly derived it from Biblical mty “men,” a frozen plural. The nineteenth century Phoenician handbooks thus naively list the form mt with the meaning “man” in their glossaries of Phoenician words. All the forger had to do to get the “Phoenician” word mt was look in these nineteenth century glossaries. The glossaries were wrong, but the forger is simply reflecting the errors of his era.

Each of Gordon’s examples is explainable in a similar way although the explanation is often even more complex and sometimes requires technical knowledge of the languages involved to understand. But it is clear that the Paraiba does not contain anything which would be unexpected in 1872.

To conclude that the Paraiba inscription is a forgery is easy. The more difficult question is to identify the forger and his motive.

My own guess is that it was done by a scholar who was either ousted from a position or passed over by Viana or Neto—in short, a case of academic enmity. The aim was perhaps to embarrass one or both of them—or to simply send them on a wild goose chase. The forger may have succeeded: It was only after several years of work that Neto repudiated the inscription.

That the forger chose this means to “get even” with his academic enemies was quite expectable in the 1870’s. Indeed, it was the season of monumental forgeries: a period of spectacular discoveries is often followed by a period of imitative forgeries.

The same thing happened following the discovery of the Dead Sea Scrolls. I myself have seen documents in Palaeo-Hebrew and in square Hebrew (properly Jewish) letters, as well as in other esoteric scripts, which were being peddled on the “scroll market” in Jerusalem—all transparent frauds.

The same thing happened in the 1870’s—after the discovery in 1870 of the spectacular Mesha Stele (or Moabite Stone, as it is sometimes called), and in 1871 of the equally spectacular Temple Stele. The Temple Stele is a fragment of the inscription from the gate of the Inner Court of Herod’s Temple, barring entry to Gentiles. As reconstructed the inscription reads: “No Gentile shall enter the fence and barrier around the Holy Place. Whoever is caught (inside) will be responsible for (his own) death.” (A second and adjoining fragment of this inscription was found in 1935.)

The Mesha Stele is a black basalt stele four feet high and two feet wide erected by Mesha, king of Moab to commemorate his successful rebellion against Israel in about 830 B.C. The incident is otherwise recorded only in the Bible itself (2 Kings 3:4ff.).

Spectacular finds like these were followed by equally spectacular forgeries. The most famous of the 1870’s and 1880’s was the so-called Shapira texts—15 strips of parchment about 3 ½ inches by 7 inches which appeared to contain a variant version of parts of Deuteronomy—in a script naturally similar to the ninth century B.C. Mesha Stele. But there were many other forgeries of the period. One of these others was the Paraiba inscription!

I would like to thank BAR editor, Hershel Shanks, for helping me “translate” my technical study into language that the lay reader can understand.

(For further details, see Frank Moore Cross, “The Phoenician Inscription from Brazil. A Nineteenth-Century Forger,” Orientalia, Vol. 37 (1968), pp. 437–460; C. H. Gordon, “The Authenticity of the Phoenician Text from Parahyba,” Orientalia, NS 37 (1968), pp. 75–80.)

Of the recurring, often bizarre attempts to find ancient Semitic inscriptions in the western hemisphere, the most prominent and frequently cited concerns the so-called Paraiba inscription from Brazil. The Paraiba inscription is said to be a Phoenician inscription carved in stone and found in northeastern Brazil in 1872. After raising a minor stir in the 1870’s, the Paraiba inscription was almost forgotten—until another copy of the inscription was publicized in the 1960’s by Professor Cyrus Gordon, then of Brandeis University and now at New York University. Based on this newly found copy, Gordon, in 1968, published an article […]