A Glossary of Lingua Franca -- Introduction

A Dead Language

What is Lingua Franca?

Lingua Franca is a pidgin, a trade language used by numerous language communities around the Mediterranean, to communicate with others whose language they did not speak. It is, in fact, the mother of all pidgins, having been in use in the Middle Ages and surviving until the nineteenth century, when it disappeared without a trace, leaving only a few anecdotal quotations in the writings of travelers or observers -- and the initial numerals in the mouths of the present-day children of Jerusalem, who use them as a counting-out rhyme, without being aware that they are not mere nonsense syllables, but the sad remnant of a once highly useful means of communication, an informal Mediterranean Esperanto. Like other pidgins, it had a limited vocabulary and a sharply circumscribed grammar, and lacked those things, such as verb tenses and case endings, that add specificity to human speech. The language was never written. No poetry, no folktales, no translation of the Bible, just a way to sell the merchandise you had to offer, or haggle for a better price on its purchase.

Observers noted that the words that constituted this pidgin were mainly of Romance origin, in particular, Italian, Spanish and Occitan, the variety of French spoken on the mediterranean coast of France and its hinterland. Non venir encora journo di sancto di vos autros? ("Hasn't your holiday arrived yet?") politely enquired one speaker of another. Note the original negative and the use of an uninflected infinitive where some kind of past tense would be needed in all "normal" languages. Apart from a few curious and tantalizing quotes of this kind, Lingua Franca seemed to be lost forever, since it died before the advent of the tape recorder or of anthropologists anxious to record a moribund form of human speech, however bizarre, and even laughable, it may have seemed.

Some years ago, in reading a charming and once-popular book in Algerian Judeo-Arabic (colloquial local Arabic of the western variety, written in cursive Rabbinic Hebrew characters) I observed that there were numerous Romance loanwords, totally absent in standard Arabic, often flanked by Arabic inflexions. Now this book was printed -- beautifully, on imperishable rag paper -- in Leghorn, Italy, a great center of Hebrew printing. During the nineteenth century Leghorn boasted sixteen Jewish publishing houses whose products are a joy to the eye. The house of Belforte, the preeminent publishing house founded in 1838, produced this Shay Lamora' ("A Gift to Him who is to be Feared" -- after Psalm 76.11. The first word is pronounced like the English word shy. This is a normalized transcription. In the transcription I describe elsewhere, it would appear as $y lmwra.) The authors of the book, however, Solomon Zarqa and Judah Darmon, lived in Oran in North Africa which was a recognized center of Lingua Franca, and it is hardly surprising that their book, directed as it was to simple, non-intellectual folk, should reflect in its language the familiarity with Lingua Franca, and frequently borrow words from the common speech. Jews were prominent participants in international trade. I published an article in Hebrew on my observations in a volume on the culture of North African Jewry (see bibliography) and later gave a lecture on my findings at an international Orientalists' conference in Toronto. The current work represents a fulfilment of my intent declared there to publish Linguae Francae Relicta, a monograph, in which would be gathered together all the fragments I could find of Lingua Franca. One need not worry too much about syntax in Pidgins, because usually it is very uncomplicated. Finding the morphemes -- themselves not great in number -- is the better part of the battle. It is interesting to note that the authors of Shay Lamora', in their preface to the first volume, note their perception of the character of their language which is

this corrupted Arabic language in which we have composed this book, so that everyone who reads it will understand immediately and painlessly. (hada lsan alerby alm$wb$ aldy emlna byh had almcHaf ba$ jmye aldy yqra fyh suby+w yfhm bla edab.)
To my mind, they had no need to be deferential to more highly regarded languages. They wrote simply and fetchingly in their native tongue, an opportunity largely denied to the present-day denizens of that part of the world, who as soon as they put pen (or printer) to paper must use Standard Arabic, a language that is essentially foreign, even though political and social circumstances require them to regard it as their own. Diglossia has its rewards, but at a certain price.

Concord and Pluralization

The main argument I made in the Toronto lecture to which I just referred in favor of a Lingua Franca origin for the loan words in the Shay Lamora' was the lack of concord between noun and adjective. (See the text of part of my lecture, here provided.) However, pluralization does occur, and it takes three forms, which correspond roughly to the two plurals, formulae and formulas, of the loanword from Latin formula.
  1. There are plurals representing the Romance plural in -s. The words wwylaj (village) and lafar (affaire) form plurals wwylajys and lafarys. The fact that the -s is enunciated suggests strongly that these words are borrowed from Occitan rather than French.
  2. There are plurals formed by adding the suffix of the Arabic sound feminine plural to the Romance morpheme. Thus the plurals of lfal+a (error) and of byzy+a (visit) are lfal+at and byzy+at.
  3. In some cases the loanword has an internal plural, known in Arabic as a broken plural. Thus kwwaj+ is to be derived from Romance cajeta. One might compare the broken plural mwa+r derived from the loanword motor, used for "autombiles" in some Arabic dialects.

In effect the plurals of the first type are the genuine Lingua Franca plurals, while the other two are the way the plurals occur when already assimilated to the Arabic language of the text of Shay Lamora'.

The Research of Marcel Cohen

Le parler arabe des juifs d'Alger (Paris, 1912) was written by the fine linguist Marcel Cohen when he was still in his twenties, and included by la Société de Linguistique de Paris in its series. (A bibliography of his life work was published in 1955 under the title Cinquante Anneés de recherches linguistiques.) Cohen clearly recognizes that the colloquial speech of these Jews contains many loanwords from Romance languages, but he is not inclined to believe that the role of Lingua Franca was significant, and prefers to hold that loans from Romance languages were distinct, and that "Provençal" was rarely borrowed.
... il apparaît que l'influence de la langue franque n'a pas été nulle sur la langage ordinaire des juifs d'Alger ... Ceci dit, il me semble pourtant prouvé par un ensembles de faits que la presque totalité des emprunts romans à Alger juif provient non de la langue franque, mais directement des diverses langues romanes. (Page 413)
In proof of this, he cites a wordlist of Lingua Franca, containing some 2000 lexical items from Marseilles dated 1830, which indicates that Lingua Franca consisted mainly of Italian words. (More on this wordlist in the next section.) Moreover, the meanings of words in Algerian Arabic corresponds better to the modern languages than to the wordlist. An typical example of a dialogue is offered here.

I find Cohen's argumentation unconvincing. The fact that the infinitive of the verb (mostly -ar, sometimes -er) is used in these items, albeit with Arabic affixes, is strongly suggestive of an underlying pidgin. And relative frequency of the four languages that are the ultimate source of the loans seems to follow the Archbishop's observation relative to Lingua Franca.

Cohen includes a number of texts which are transcribed from oral materials he obtained from informants, mostly on the customs of these Sephardic Jews, but a commercial item, which is apparently transcribed from Hebrew script, is of particular interest to us. (Page 514.)

He observes that it is an instance of the langage commercial écrit, rempli de mots étrangers désignant presque tous des opérations ou des institutions commerciales. The letter, in which the writer complains that his correspondent is not acting appropriately, uses loan words in exactly the same way as Shay Lamora', Romance infinitives being squeezed into Arabic paradigms (his necessary diacritics are not reproduced here):

All of this (and more) in a letter nine lines long! Cohen says that "almost all the words [the foreign words] are French," but the truth is that all the words are Lingua Franca -- witness prisantar, aksiptar, rifuzar, which all have analogues in French, but are not French. In the other texts he cites, which were presumably first given orally by informants, and then transcribed, both in his scientific transcription and in the Hebrew script used for Judeo-Arabic.

I am including Cohen's valuable data in this Glossary, but have so far concentrated on items I extracted from Shay Lamora', since Cohen's items are readily available in his book, which is out-of-print, of course, but may be found in many research libraries, that of Brown University in Providence, Rhode Island, for instance.

La dictionnaire de la langue franque ou petit mauresque

The wordlist cited by Cohen was published in 1830. In 1828, doubtless under provocation, the Dey of Algiers had impetuously struck the French ambassador, M. Deval, with his fly-whisk. This insult to the honor of France was avenged in 1830 by an invasion, which resulted in 1842 in the total annexation of Algeria. This story ended only in 1962, when Algeria achieved independence after an eight-year war in which more than a million people lost their lives. The wordlist is clearly meant to assist French-speaking individuals who were moving to Algeria. It is French to Lingua Franca only, and the transcription follows the conventions of French orthography, much as present-day tourist guides will offer a mimicked rendition of foreign words. Accents are not used to indicate stress, but rather to ensure that French speakers will pronounce the vowel. Accordingly, the wordlist is a practical guide, rather than a scientific description. I have abandonned his inconsistent transliterations, using rather the standard spelling of the words of origin, noting the rather frequent elision of a final consonant. These words are offered in a separate section of the Glossary; many of them are, of course, non-standard. Sp. ayudar, for example, will appear as ajudar, the j pronounced as in English, a pronunciation heard not infrequently in the south of Spain. Verbs are recognizable by the -r ending, "to" is omitted in the English equivalent. Verbs of Italian origin routinely have the final vowel elided, thus making them resemble Spanish verbs. (E.g. ragion "motive", ragionar "to reason." It is quite probable that if the author did not know the Lingua Franca word for a particular French word, he threw in the Italian equivalent, and travelers doubtless did this in any case. It may be noted that the book was published in Marseilles, which is quite close to the Italian-speaking area. Another point to note is that Arabic lacks the u/o distinction that is so common in Western European languages, and these phones are in effect allophones. Hence the European ear would frequently here the o sound as u in the mouths of Arabic speakers. (And vice versa; foreign reporters commented on Iraqi references to President "Bosh".) Lingua Franca clearly never achieved any kind of standard form.


Maltese is a Semitic language closely related to Western Arabic with a large Romance vocabulary. So far as I can determine, these loans are from Sicilian, and do not display the characteristic features of Lingua Franca, such as fossilized verb infinitives.

A Dying Language

Judeo-Arabic has existed as a distinct entity since medieval times, and still exists, albeit barely. It contains some world-class classics; the great Maimonides' Guide for the Perplexed was written in this language. This text was studied in the Philosophy department at my university not long ago in an English translation, and I took a course in the original Arabic text (the only student for a full year!) with Professor Pesah Shinar at Dropsie College in the early sixties. The famous libraries of the world are full of manuscripts in Judeo-Arabic which have never been edited, although many of them would merit scholarly concern. And there is a large number of printed books in the later, local, colloquial Judeo-Arabic also. (See Vassel in the bibliography.)

Messaud Maarek

Messaud Maarek was a Jewish Tunisian journalist at the end of the nineteenth century who made a brave attempt to create a literary medium of Tunisian Judeo-Arabic. Clearly, he was an educated man familiar with standard Arabic, and he tried to make the language he used in his literary efforts accept some of the norms of the standard language, without eliminating some of the firmly entrenched characteristics of Western Arabic, such as the first person of the imperfect tense of verbs in n- with n- + -u for the plural (n$uf/n$ufu, for example for I/we shall see.) Maarek had two strikes against him. First, he was himself no Shalom Aleikhem or Isaac Bashevis Singer. No first-rank, gifted writer arose in modern Judeo-Arabic to match the luminaries of Yiddish literature. The attempts to engage in belle-lettres in Judeo-Arabic literature are disappointing, to say the least. Second, French had acquired such a pervasive, mystical status of being the true, God-given language of culture, recognized from Moscow to Rabat, if not further, that no one was likely to cultivate an unfairly despised Jewish dialect, when so many opportunities arose to learn to express one's self in the charming, universal French language, which connected the speaker with a vast francophone world.

Judeo-Arabic v. Yiddish

Judeo-Arabic is comparable to Yiddish in that it is a language exclusively in use among Jews, written in Hebrew characters, with diacritics -- which never became standardized -- to represent the sounds of Arabic which Hebrew does not possess. Of course, the standard Arabic alphabet itself developed diacritics to cover the absences of certain phonemes in the underlying Aramaic script. It was different, however, in that it was most often spoken among peoples who also spoke Arabic, albeit in a differing dialect. A Baghdadi placed in a dark room with other Baghdadis could immediately detect if an interlocutor was Muslim, Christian or Jew, since the community dialects were distinctive, although similar. This phenomenon of diglossia was researched by Ferguson, and evoked much interest on account of its broad implications for sociolinguistics. (See Ferguson in the bibliography.)

The view of Wexler

Paul Wexler has brought Yiddish closer to the Arabic model by proposing that Yiddish is in reality a Slavic language, albeit with a principally Germanic vocabulary, like Sorbian. (See Wexler in the bibliography.) Despite the controversial nature of his proposal (which drags in Modern Hebrew too!) I believe that Wexler's views are fundamentally correct. However, rather than argue whether Yiddish is or is not a Slavic language, we should come to the conclusion that the standard "family tree" model of language descent is fundamentally flawed. Languages are not trees, and they are not animate creatures. One might claim jokingly that their origins are a kind of hypertext, with links to all kinds of other forms of speech, and not traceable back to some supposed ur-language. In trying to do this, we get into an Escher-like progression which takes you back whence you came.

The Future of Judeo-Arabic

What is the future of Judeo-Arabic? In all probability it has none. It is dying fast, and will soon be gone for ever. Yiddish may experience a substantial revival. It is favored at some western universities and in Israel. Even Judeo-Spanish may have a limited survival, although that too seems unlikely. The ancient Irish language is making some sons of Erin feel good, and the decline in the number of speakers, encouraged by those fine Englishmen who prohibited the Irish to speak their own language, seems to have been arrested. Judeo-Arabic is in a particularly unfortunate category, however. When I was doing some research on Judeo-Arabic in Israel in 1970, I had occasion to interview a Jew from Libya, and encourage him to give me some samples of his speech. His lively brood of children found this totally hilarious. They had never before heard these strange vocables emerging from their father's mouth. Their father told me that he deliberately avoided speaking Arabic, in order to identify with the living Jewish culture of Israel. For him, his native tongue was a symbol of an oppressive life of exile which he preferred to forget. Mr. Gadafi might assert that his beloved Jewish former citizens were welcome to return to Libya, but this citizen was not having any of that. He belonged where he was now, not where he had been born.

Judeo-Arabic, then, is reaching the end of the road. But it leaves behind a distinguished history of medieval philosophical writings, folk literature, poetry, even modern newspapers which once flourished. And -- in my humble opinion -- a startling clue to the lost Lingua Franca language, which, as the Archbishop of Palermo declared in the seventeenth century, was once known in every house in the great cities a short hop away across the blue Mediterranean.

A Living Web

Since I wrote my remarks in Hebrew less than a decade ago, a much better option for encapsulating my findings than a monograph has presented itself. I would call this a cybergraph, a dynamic collection of information quite unlike the frozen monograph, which cuts off one's researches at a given point in time, after which a new edition may be contemplated -- contemplated but probably not brought to fruition in today's environment when printing costs are so high, and distribution of limited-circulation scholarly tomes is quite difficult. I can publish my findings in their current state, and update them weekly, or at whatever interval I find convenient, a truly awesome prospect which is the direct result of the innovative initiatives of those who formed first the Internet, and then the World Wide Web. I face this prospect with a sense of awe, and a great deal of excitement. Already I have found that, with the help of those astonishing search engines, such as Yahoo! and WebCrawler, my writings get much more exposure on the Web than I could ever achieve in cold print. The need to find a publisher has disappeared. I wish the publishers well, and recognize their genuine contributions to the dissemination of scholarship, but I am not averse to being exempt from dealing with them in this instance, despite the pleasure one feels at holding in one's hand a book with one's name on it in large letters. True, an effort such as this might be considered the work of my personal vanity press, but my books and articles have passed muster in the past, and I hope this work will meet with a favorable reception. To help meet this problem, if indeed it is a problem, I have solicited some approbations in the spirit of the haskamot prepended to works of rabbinic scholarship which testify to the appropriateness of the work and the author. This may replace the kind of monitoring that submission to a scholarly publishing house normally brings forth. Whether favorable or not, your opinion can find its way to me very easily by electronic mail at corre@csd.uwm.edu, and I hope you will add information or criticize, to the betterment of the product. In accordance with the free spirit of the Internet, I choose not to copyright these materials. You may reproduce or disseminate them, although I would prefer that you keep them in electronic form, and not kill trees. Maybe we are reaching the time indicated in Psalm 96.12 when all the trees of the forest shall shout for joy that they can live their natural lives and not feed our voracious appetite for paper. I also request that you give the source where appropriate:
Having read this introduction and the dependent materials, you can in future access the glossary by the shorter URL
Links are provided to other areas of this cybergraph.

Text and Transcription

A Sample Text

A brief sample text is offered. You will see that the Lingua Franca words there stick out like raisins in a bun, even if the aberrant Arabic of the text should give some difficulty to read. In my initial Hebrew essay on Loanwords in the Shay Lamora' ("milim sheulot besefer shay lamora'") I had the luxury of being able to cite in the Hebrew script, simply substituting the square script for the rabbinic cursive of the original. But this greatly limits the audience for these findings, since they may be of interest to sociologists, anthropologists, linguists or historians who are not too familiar with the beautiful Hebrew alphabet with its entirely sensible single-case letter system -- no caps, or maybe no smalls, who knows?

The Problems of Transcription

Finding a one-to-one transcription of Judeo-Arabic of whatever ilk to Roman script is no great problem, provided you are not interested in knowing how the language sounded, since vowels are scantily represented. Here, however there is the problem of the limitations of the HTML character set. Transcriptions of Semitic frequently use diacritics such as a dot under a letter to indicate a sound not found in European languages, but these are not available in the standard ISO HTML charset. Accordingly I press the plus sign (+) into service to represent the emphatic t of Semitic which would normally be represented by a dot under the non-emphatic t, and a capital H to represent the eighth letter of the Hebrew alphabet, compared with h which represents the fifth letter. The beauty of the Internet is that if at some time in the future better transcriptions become available, it should be fairly easy to make some global changes which will exploit them. Take a look at the sample text to see how these problems are handled.


The study that I offer here is an instance of a literary feature I would call embedding. The recognition that the text of Scripture has poetic fragments embedded in it is an important help in understanding the text at a deeper level. Genesis 4.23-24 is an example of such embedding. In the Authorized (King James') version of the Bible such embedding is not apparent in the text, but all modern Bible translations routinely print such items in an indented fashion, and usually point out its character (see Metzger in the bibliography). Metzger describes this example in his note as "... an ancient song, probably once sung in praise of Lamech... here quoted..." This simple recognition is very helpful to the thoughtful reader. The First Book of Maccabees in the Apocrypha is an even more striking example of this phenomenon. Metzger annotates I.24-28 as a "fragment of a contemporary poem" and there are other examples throughout the book (I.36-40, 2.7-13 et passim.)

A much more controversial example is William H. Shurr's claimed discovery of new poems by Emily Dickinson embedded in her correspondence (see Shurr in the bibliography). The blurb declares that "Although many critics have commented on the poetic quality of Dickinson's letters, William Shurr is the first to draw fully developed poems from them." When this book appeared, a highly critical review appeared in the Chicago Tribune newspaper, asserting that this so-called discovery is merely the product of Shurr's fevered imagination. Some credence is given to this stricture by Shurr's assertion in the preface that "We [i.e. Shurr and his two assistants] have all studied the letters, and each of us has found poems that the others missed." "No wonder!" the cynic will declare, but I feel that Shurr is really onto something, although one may question individual items that he may have included in order, perhaps unconsciously, to achieve a critical mass. When poets write prose, they cannot leave their muse behind completely. I get the same feeling when reading Amichai's enchanting modern Hebrew prose, but feel no compulsion to dig out therefrom a new corpus of his poetic work. I realise that by citing this controversial example I risk transgressing the Talmudic precept never to give an "opening of the mouth to the accuser" (Babylonian Talmud, Berakhot 19a) -- in the form of Satan, or an unhappy reviewer -- but honesty requires me to do so.

At the other end of the probability scale is Samuel Miklos Stern's stunning discovery of the jaryas "exits", embedded in the envois of medieval Arabic and Hebrew poetry which solved conclusively the puzzle of these previously incomprehensible lines, and added a new chapter to the history of the Spanish language. (See Stern and Latham in the bibliography.)

I have used this discovery principle of embedding in a more strictly linguistic context, i.e. finding the lost vocabulary of the Lingua Franca principally within colloquial Judeo-Arabic texts. The reader will, I trust, find that the probability that my hypothesis is correct is greater than fifty percent, but that is for him or her to decide.

Allow me to interject a personal note. When I was in my early twenties, I was much inspired by the example of Samuel Stern whom I mentioned above. Dr. A. Altman, who collaborated with Stern on some scholarly projects, gave me the job of editing and proof-reading one of Stern's articles which Altman was preparing for publication, and I made a memorable visit to Stern in the house of his friend Walzer in Oxford. Stern was an awesome scholar, and his tragic death from asthma before he reached the age of fifty deprived us of more insights from his deep scholarship. But he had already done more than the equivalent of a life's work in his all too brief years. He is much in my mind as I dig out these little nuggets of Lingua Franca from the mine of colloquial Judeo-Arabic, and I like to think my work is a minor analogue of his digging out old popular Spanish from the mine of formal, highly literary, Hebrew and Arabic poetry. Would that I could submit to him this research, for he would give a generous yet unflinching assessment. But he already sits on the front row in the yeshiva shel ma`la, the Academy on High. I hope his view of my efforts is favorable, because he of all people would perceive the truth.

Go on to the Glossary
Go back to Title Page
Go back to Home Page
Alan D. Corré