1. Introduction
  2. Some rambling thoughts
  3. A stratigraphy of Maltese surnames
  4. Hull’s theory of a Girgenti colony
  5. Early census taking and surname rankings
  6. Census 2005 and the commonest surnames in Malta
  7. The pantheon of Maltese surnames
  8. Surname frequencies by location
  9. A snapshot of Gozo
  10. The ‘Australian’ parallel sample
  11. Cognates and doublets
  12. Multiple surnames
  13. Extinct surnames

Cognates and doublets

Mario CASSAR

Should Lombardo and Lombardi be analyzed as the same surname? What about De Marco and Demarco, German and Germani, Laganà and Lagana, Manicolo and Mannicolo, Xerri and Scerri? The answer is not that simple as in all cases one has to deliberate on orthographic, morphological, phonological, and semantic factors. However, some basic criteria are on hand to mitigate the problem.

Allographs are variant forms of a grapheme in a way that in writing they are seen as analogous to a phoneme. Such is the case of x, which in Sicilian and southern Italian is often the medieval or archaic rendering of sc. It also happens to be the present Maltese grapheme with the same phonetic value /∫/. Hence there is no point doubting that Grixti/Griscti, Axiak/Asciak, Xerri/Scerri, Xicluna/Scicluna, Xiberras/Sciberras, X(i)riha/Sc(i)riha are exactly the same surnames and should be treated as such when compiling statistical data on the local cognominal pool. It is important to note that Census 2005 has treated all pairs and bundles discussed in this section as separate surnames.

Surnames are sometimes agglutinated with prepositions signifying paternity or provenance. Hence De Gaetano/Degaetano, De Marco/Demarco, De Bono/Debono, D’Emanuele/Demanuele, De Martino/Demartino, De Noto/Denoto, and the like are nothing but cognates. The same principle applies to La Rosa/Larosa and Sant’ Angelo/Santangelo. The differences constitute only arbitrary orthographic conventions. Some surnames which ought to show a final accented vowel often do away with this diacritic: Laganà/Lagana, Ciarlò/Ciarlo, Faurè/Faure, Burlò/Burlo, Marmarà/Marmara, Gialanzè/Gialanze. We are again dealing with cognates.

The following bundles should also be seen as reflecting the same surname as only orthographic idiosyncrasies seem to be in play: Meilak/Meilaq/Meylak/ Meylaq/Mejlak/Mejlaq; Musù/Mousù; Theuma/Teuma/Thewma/Tewma; Briguglio/Brigulio; Calleja/Calleya/Kalleya; Mugliett/Mugliette/ Muliet/ Muliett/ Muliette. De and Di are cognates; hence, both De Pasquale/Di Pasquale and De Battista/Di Battista should be treated as one surname.

Differentiation should start with morphological considerations. Surnames with different endings, particularly those representing singular forms in -o/-e and plural forms in -i should be treated as different surnames. This category is quite copious: Lupo/i, Mariano/i, Corrado/i, Costanzo/i, Rosso/i, Albano/i, Veneziano/i, Lombardo/i, Magro/i; Conte/i, Forte/i. The same criterion applies to surnames differing in their gender marker: Gravino/a, Giardino/a, Navarro/a.

Surnames in full form should be set apart from their cognate apocopated forms: German/Germani, Gambin/Gambina, Paris/Parisi, Palmier/Palmieri. Surnames with patronymic suffix De/D’ should be distinguished from their counterparts which have lost the preposition: Giorgio/Degiorgio, Felice/De Felice, Alfonso/D’Alfonso, Flavia/De Flavia, Giglio/De Giglio, Arrigo/D’Arrigo, Alessandro/ D’Alessandro, Esposito/D’Esposito, Anastasi/D’Anastasi.

Voiced/devoiced cognates (e.g. Giliberti/Ciliberti, Montalto/Montaldo) and dialect variations should be counted as separate surnames (e.g. Aloisio/Aloiso, Barberi/Barbieri, Gialanze/Gelanze Bongiovanni/Buongiovanni, Dandria/ D’Andrea, Seracino/Saraceno). Traditional surname Abdilla should be differentiated from Abdalla(h) and Abdullah which are obvious recent additions from North Africa. Grech/Greco and Barun/Barone, despite being doublets, must be differentiated as well. Doublets are two words in a language which are historically from the same source, but with different intervening stages.

It is much harder to be specific when dealing with geminated and metaphonic forms. The first group includes: Seracino/Serracino, Manicolo/Mannicolo, Bonnici/Bonici, Carabot/Carabott, Azzopardi/Azzopardi/Azopardi, Cacciattolo/ Cacciatolo, Benetti/Bennetti, Baluci/Ballucci, Sammut/Samut, D’Emanuele/ D’Emmanuele, and Selvagi/Selvaggi. The second group includes: Sevasta/Savasta, Salomone/Salamone, Pantalleresco/Pantallaresco/Pantelleresco, Sciberras/ Sceberras, Cacciattolo/Cacciottolo, Felice/Felici, Miruzzi/Meruzzi.

Caution is always recommended, and one should never dogmatize. Even these guidelines are susceptible to criticism. The conundrum intensifies when semantic differences seep through. Is Muliett(e) really the same surname as Mugliett(e). The former is probably an apocopated form of Sicilian mulièttu (sometimes muliètta), cognate form of mulettu, a generic term denoting various species of mullet. The latter can be a mere respelling, but alternatively it may be related to the Muglia family. Baluci and Ballucci may truly be variants, but the former can be a dissimilated form of Balducci, while the latter can be another form of surname Baluce.

In some cases variants are neither genuine cognates or doublets but mere scribal errors. In fact misspellings have also to be taken into consideration. Aguis, Carauna, Aqulina, and Busutill obviously stand for Agius, Caruana, Aquilina, and Busuttil.

The same problems resurface when dealing with surnames which are neither Italian nor Sicilian. British cognominal variants in Malta abound: Cox/Cocks, Burke/Bourke, Griffiths/Griffits, Giles/Gilles, Payne/Paine(e), Reed/Read, Swaine/Swaine, Haig/Haigh, BonnettBonnet/Bonett, Gracey/Gracie, Dixon/Dickson, O’Brian/O’Brien, Stroud/Strout. These should be treated as separate surnames. So should Connor/O’Connor, Dougall/Mac Dougall, Cann/McCann; Edward/Edwards, Field/Fields, Owen/Owens. However, only trivial orthographic convention seems to set apart McLean/Mc Lean, McNamara/Mc Namara, Mackay/Mc Kay/Mckay/Mac Kay/McKay.

French and Spanish cognates include: Guillaumier/Gullaumier, Marquett/Marquette, Critien/Chretien/Chritien, Mompalau/Monpalau, Gomez/Gomes, Garcia/Garzia (Garnier de terra Bengü en France ?) Dorel, Poussieghes, De la Roche.

The Salves have Ivanov for males and Ivanova for females. This artificial delineation has to be neutralized when calculating the number of extant surnames.

Census 2005 claimed that there are some 12,310 different surnames in Malta. This number has to be trimmed down considerably. First of all, one has to subtract the particularly high number of double surnames (3,882), (74) as all Mifsud Bonnicis and Borg Costanzis, in real fact, are nothing but Mifsuds and Borgs. This practice may be important for bureaucratic purposes and obviously for the bearers themselves, but from the anthroponymic point of view only the first surname counts. That would leave 8,428 surnames. Then all the clear-cut allotropic surnames discussed above have to be eliminated as well. This pruning exercise will reduce the actual number to around 7,000.

Notes :
  1. NSO press release (5/6/06), unofficial count.

Published with permission from Mario CASSAR


  1. Introduction
  2. Some rambling thoughts
  3. A stratigraphy of Maltese surnames
  4. Hull’s theory of a Girgenti colony
  5. Early census taking and surname rankings
  6. Census 2005 and the commonest surnames in Malta
  7. The pantheon of Maltese surnames
  8. Surname frequencies by location
  9. A snapshot of Gozo
  10. The ‘Australian’ parallel sample
  11. Cognates and doublets
  12. Multiple surnames
  13. Extinct surnames
Top