Perhaps, there is nothing that distinguishes humanity from the rest of the animal kingdom more than the idea of individual identity as exemplified by conferral of a name.  Concomitantly, it may be assumed with reasonable justification that human language is a fundamental prerequiste for this practice.  That is not to say that other species are not consciously self-aware and, perhaps, may even possess some form of sentient identity (anyone living with a beloved dog or cat will likely assert as much with considerable warmth).  Indeed, in the case of cetaceans (i.e., whales and dolphins) some advocates have argued that they do have language and even self-conferred names, but this proposition remains debatable and, as yet, is not generally accepted.  Likewise, in a controlled laboratory environment great apes, such as chimpanzees and gorillas, have been successfully taught Amercian Sign Language as well as other types of symbol manipulation, which would seem to indicate some innate aptitude for langauge acquistion (typically similar in complexity to, perhaps, that of an average two-year-old human child).  Moreover, it is likely that some form of this behavior also occurs in the wild and, accordingly, has a survival value for the species.  Nevertheless, although apes exhibit complex social structures that evidently recognize individuals and even exhibit rudimentary culture (as shown in the pioneering work of Dr. Jane Goodall among others), there is no evidence that they use names in anything like the same sense that humans do.  Therefore, it seems safe to conclude that the custom of naming is unique to humanity and probably emerged in remote prehistoric antiquity concurrently with the ability for complex language.  Concomitantly the formal study of names and naming practices is called onomastics (or onomatology) and, as such, should be regarded as a specialized branch of anthropology and linguistics.  (In particular, anthroponomastics concerns personal names and toponomastics concerns place names.)

Personal Names

     Obviously, in the most fundamental sense a personal name is merely an abstract linguistic symbol for an individual person.  Moreover, at birth human infants are unable to appropriate names for themselves, consequently, another older person must confer a birth name.  Thus, it would seem natural and, perhaps, beyond doubt that this would be done in the great majority of human societies by one or the other of the child's parents.  Alternatively, some other relative or a clan leader might also have had this formal responsibility.  In any case, it would seem probable that beyond being a simple identifier, the chosen name would have a significant meaning that might serve several purposes, viz., to invoke some blessing from supernatural powers, to identify the new individual with ancestors, to describe some putative trait of the child, to mark an event associated with the child's birth, etc.  Of course, before the advent of written language personal names were by necessity only spoken and, therefore, might be chosen purely for euphony.  Indeed, choice of a name is intensely personal and, as such, defies any simple categorization.  Furthermore, it seems that in contrast to modern Western societies in which with a very few exceptions, birth names remain fixed throughout a particular individual's lifetime, in more traditional cultures an person typically might change names at various stages of life.  Such customs were observed to be common among Native Americans at the time of European settlement and persist among other aboriginal cultures even at the present day.  Another aspect of a personal name in such societies is its totemic power, that is to say, one who invokes a person's name might be able to gain some kind of "power" over that person, i.e., as a form of sympathetic magic.  Clearly, to deny such power from human enemies or even malevolent spirits, it would be desirable for one to keep his or her "true" name secret.  Accordingly, one would have a nickname or "public" name distinct from any secret true name.  Indeed, such customs have been practiced at various times and places among various social and ethnic groups.
     Upon the advent of writing, civil records began to be kept and formal histories composed, which would become very confused and complicated if personal names were changed frequently.  Therefore, it is probable that personal names tended to become fixed at the advent of written language, as they presently remain.  Indeed, this seems to be true for the great majority if not all literate cultures.  Even so, naming practices were (and are) highly variable and defy any easy generalization.  In any case, in prehistory and early antiquity, as well as the early medieval period, a mononym, i.e, a single proper name, was likely sufficient for all but the most influential members of society.  This name was likely chosen for a variety of reasons as asserted above.  Concomitantly, in the Christian era it became customary to confer a personal name at baptism, which then would be one's Christian name.  Naturally, these were customarily chosen from among canonized saints or the holy angels (of which only three are named in the Bible, viz., Michael, Gabriel, and Raphael) and were given both to males and females.  Of course, the fundamental purpose of the Christian name is to identify an individual as an adherent to Christianity and by extension as a member of the universal church.  Furthermore, in some ethnic communities given names and Christian names would have been one and the same, but in others they would have been different.  Similarly, individuals that became members of monastic communities commonly took different names to signify that they had left their old lives behind.

Roman Naming Practices

    The Romans originated a system of nomenclature consisting of a combination of personal and family names.  Accordingly, Roman grammarians regarded the combination of praenomen, nomen, and cognomen as essential elements of a Roman name and designated this form as the tria nomina (the three names).  As a matter of history, Roman naming conventions were not fixed, but evolved beginning as early as the seventh century B. C. and continuing until the end of the seventh century A. D.  Even so, although usage of Roman naming conventions evidently disappeared during the early Middle Ages, they, nevertheless, substantially influenced later European naming conventions, many of which survive in modern cultural practices.  Moreover, names conferred according to the Roman system may be regarded as a distinguishing feature of Roman civilization.  As might be expected, the most important element of the tria nomina was the nomen gentilicium, i.e., the nomen, which was a hereditary name (or surname) identifying a person as a member of a distinct gens, i.e., family, tribe, or clan, which constituted an extended Roman family, all members of which shared the same nomen and claimed descent from a common ancestor.  Particularly in the early Republic, the gens observed its own sacred rites and established private laws, which were binding on all members, although not on the Roman community as a whole.  In written form, the nomen was usually followed by a filiation (or patronymic), which was derived from the personal name of a particular individual's father (or more rarely, the name of the mother (a matronymic) or other antecedents).  The nomen was preceded by the praenomen, i.e., a forename or first name, which was a personal name that distinguished between different members of the same family. The praenomen was chosen by a child's parents and bestowed on the dies lustricius (or "day of lustration"), a ritual purification performed on the eighth day after the birth of a girl or the ninth day after the birth of a boy.  Generally, all of the children in a family would have different praenomina and although there was no law restricting the use of particular praenomina, the choice was usually determined by custom and family tradition.  Typically, an eldest son was named after his father and younger sons after their father's brothers or other worthy male ancestors.  Consequently, the same praenomina were passed down in a family from one generation to the next.  Not only did this emphasize the continuity of a family through many generations, but the selection of praenomina also distinguished the customs of one gens from another.  Indeed, patrician gentes limited usage of praenomina much more than did the plebeians, which reinforced the exclusiveness of patrician social status.  Nevertheless, although the tria nomia was recognized throughout most of Roman history, usage varied and all three names were not required or used at various times.  Indeed, during the period of the Roman Republic, only two essential elements of a name were required, viz., the praenomen and nomen.  Even so, the cognomen first appeared among Roman aristocracy as early as the sixth century B. C., i.e., at the very beginning of the Republic, but was not widely used among plebeians until the second century B. C.  Even then, not all Roman citizens had cognomina and until the end of the Republic, the cognomen was regarded as less significant and non-official.  Within this context, in the later Roman Republic, the praenomen and nomen were frequently followed by the name of a citizen's voting tribe (as a type of cognomen).  Additional, cognomina could be either personal or hereditary, or a combination of both.  Accordingly, the tria nomina was commonly expanded to include four or more names, e.g., "Quintus Fabius Maximus Verrucosus Cunctator", who was in the thrid century B. C., a notable Roman politician and general.  In contrast, during the Empire the cognomen became the principal distinguishing name and although praenomina never completely vanished, the essential elements of a Roman name from the second century A. D. onward were the nomen and cognomen.  Moreover, in the later Empire, aristocratic individulas used several different schemes for assuming or inheriting nomina and cognomina, both to signify rank and to indicate familial and social relationships.  Consequently, some Romans came to be known by alternative names or signia, but at this time the full nomenclature of most Romans, even among the aristocracy, was seldom recorded.  For a variety of reasons, the Roman nomenclature system decayed following the collapse of the Western Empire. The praenomen had already become rare in written sources by the fourth century and by the fifth century it was retained only by the most conservative elements of the old Roman aristocracy.  During the sixth century, as Roman institutions and social structures gradually disappeared, it became unnecessary to distinguish between nomina and cognomina and by the end of the seventh century, the majority of people in Italy and western Europe had reverted to the ancient practice of single names.  Nonetheless, many of these mononyms originated as adaptations of Roman names and, as such, have survived into modern times.
    Of course, Roman society was extremely patriarchal and naming conventions were essentially masculine.  Even so, Roman women originally shared the same two name convention as men; but as time progressed the praenomen became less necessary to distinguish females; hence, feminine praenomina were gradually discarded or replaced by informal names.  Accordingly, by the end of the Republic, the majority of Roman women either did not have or did not use praenomina and were called by their nomen alone or by a combination of nomen and cognomen.  Praenomina could still be given when necessary and as with masculine praenomina the practice survived well into the Empire, but proliferation of personal cognomina eventually rendered feminine praenomina entirely obsolete.  Therefore, the period during which the majority of Roman citizens were designated by exactly three names was relatively brief; however, because most of the important individuals living during the most completely recorded period of Roman history, viz., the first century B. C. and the first and second centuries A. D., were distinguished by all three names, the tria nomina remains the most familiar conception of a Roman name.  Accordingly, in modern Western society personal names (typically first names) as they presently exist, would seem to descend from the classical Roman praenomen.  Likewise, surnames can be regarded as a modern adaptation of the nomen.  There is no exact equivalent of the cognomen, but a nickname would seem to be the closest analog.  The modern middle name evidently has no Roman analog, unless it is used instead of the first name as the primary designation of an individual, in which case it can be regarded as corresponding to the praenomen.

Modern Surnames

    As indicated above, by the medieval period most people in Europe were identified simply by a mononym or single proper name.  (The only exceptions occurred among the nobility.)  However, in the later Middle Ages it became necessary to distinguish individuals having the same proper name by some kind of additional non-hereditary informal name or byname.  These were appended to a proper name, therefore, resembling a Roman cognomen, and are generally of four different types, viz., patronymics or matronymics, locative, occupational or social, and characteristic.   These four types exist in almost every European language, but frequency and grammatical construction are quite variable.  Moreover, for simplicity, in this work only formation of English bynames will be considered., but similar conventions apply to other European languages.
    Of course, a patronymic identified an individual as the son (or daughter) of a particular father.  More rarely a matronymic identified an individual as the son (or daughter) of a particular mother.  As might be expected, patronymic naming was much more common than matronymic naming and is a very ancient practice.  Accordingly, there are three main patronymic forms in English.  Naturally, the earliest and simplest English form is the unmarked patronymic, which merely appends the parental name to the child's proper name, e.g., in 1209, one finds "Geoffrey Anketil" identified as the "son of Anketil".  Later marked patronymic forms appeared, one of which, similar to the possessive form, merely appended "s" or "es" to a father's proper name, e.g., in 1327, "Thomas Richardes" literally meant "Richard's Thomas"; and one that appended "son" to a father's name, e.g., in 1324, "Robert Willeson" corresponds to "Will(e)'s son".  Clearly, these the two marked forms descend into modern English as "Richards" and "Wilson".  Moreover, the third form is typically constructed (as in this example) using a diminuative form of a popular name in the Middle Ages and not using the corresponding full form, perhaps, descending in modified form from Old English.
    Obviously, a locative byname identifies an individual by the place where the person lived, worked, or was born.  Accordinly, locative forms can be classified into two categories, viz., toponymic and topographic.  Of these two, a toponymic byname derives from a named locality, perhaps, a village, town, county, or even some named natural feature and, therefore, the toponymic form invarably incorporates a proper noun.  In Middle English (which was heavily influenced by French after the Norman Conquest), this form typically appears with a place name preceded by the French preposition "de", which is equivalent to English "of" or "from".  Hence, one finds examples such as "de Brunnesley" in 1198 and "de York" in 1324.  Of course, toponymic bynames can also refer to a nation, kingdom, province, etc.  Since these were large civil divisions, a toponymic byname derived from such a source, e.g., Wales, would be ineffective to distinguish an individual actually still living in such as division.  Therefore, this kind of toponymic byname arose to identify individuals that had moved a substantial distance from their original residence (presumably, to some other kingdom or province).  Indeed, the byname "Simon de Welsche" would not have uniquely distinguished a "Simon of Wales" actually living in Wales, but in Bedfordshire where Welshmen were not common, it could suffice.  Subsequently, in the fifteenth century usage of the prepositonal form became obsolete and the preposition simply disappeared.  (Even so, forms without the preposition can be found in earlier records such as "Richard Wangeford" in 1296.)  In contrast, topographic bynames do not include proper nouns and are derived as a concise description of a local place or natural geographical feature.  Therefore, a topographic byname might be regarded as a sort of primitive address.  This form also included usage of a preposition, but English instead of French, and which by far the most common was "at", and which was generally combined with the definite article "the" as in "atte" literally meaning "'at the" (or some minor variation).  Typical examples are found as "Attemille" in 1242, literally meaning "at the mill"; similarly in 1243 "Attewode" can be interpreted as "at the wood".  Of course, other prepositions also occur, e.g., in 1194 "Vnderegge" meaning "under edge", thus, denoting someone who lived at the foot of an escarpment.  Likewise, "Overthebek" which, clearly is to be interpreted as "over the beck", viz., a brook or stream; in 1243, "Bithewaye" meaning "by the way", denoting someone who lived by the road; and in 1333, "Theffelde", which, obviously, means "in the field".  All of these provide good examples of topographic bynames.  Moreover, perhaps the least common preposition used is "of', but although it appears rarely, in 1327 an example is peovided by "Othe felde" which, of course, means "of the field".  As might be expected, in the centuries following 1400, the preposition (and article, if present) was eventually lost.  Even so, in some cases they survived and became fused with the noun, as in the modern surnames, "Atwood", "Atwater", and "Underhill", among many others.
    Naturally, occupational and social bynames identify an individual by occupation or social status (which in the Middle Ages were closely linked, much as the Indian caste system that still survives down to the present).  An obvious example is provided by "John Smith", denoting, typically, an indvidual named John that was a blacksmith (although other types of smiths could also be implied, viz., a goldsmith, silversmith, etc.; however, in these cases, the compleete occupational designation usually appears, e.g., "John Goldsmith").  Again, before the fifteenth century occupational bynames are most frequently found in written usage preceded by a French article, viz., "le" (masculine) or "la" (feminine), but this may not represent spoken usage.  Typical examples are provided by:  "le Bakere", which in 1177, of course, denoted a woman as "the baker"; "la Lauendere", which in 1253 denoted a woman as "the laundress"; in 1275, "le Fithelare", literally denoting a man as "the fiddler"; and in 1279, "le Horsmongere", obviously denotes a man as "the horse-dealer'.  Bynames referring to rank and social status also appear with an article, such as in 1221, 'le Freman", or "the freeman"; and in 1255, "le Erl", or "the earl".  Even so, the article did not always appear even in early records; hence one finds "Henry Pope" in 1296 and in the thirteenth century a number of individuals are found with the byname "Kyng".  Concomitantly, locative and social bynames overlap in ethnic names.  In this case, an individual is identfied by nationality, religion, or ethnicity.  Moreover, it must be noted that in the Middle Ages, national groups may be associated with much smaller geographical regions than the nation-states of the modern world.  Accordingly, in 1230 "le Bret" denoted a Breton or Briton man from Brittany in northwestern France.  Likewise, in 1276, "le Pycard" denoted a man from Picardy in northwestern France.
    The fourth type of informal name is called a characteristic byname (or nickname) and, as such, encompasses all instances that are inconsistent with any of the first three types.  In this case, usage of French articles was arbitrary and sporadic, but nevertheless, again, articles essentially disappeared after 1400.  Within this context, some common characteristic subtypes can be identified as denoting physical, mental, or moral characteristics of a particular individual.  Indeed, although characteristic bynames generally appear to be complementary, they may have been ironic or even degoratory.  For example, in 1177, "Henry Bigge" might have been a small rather than large man.  Naturally, the most common nicknames were quite simple and concrete:  in 1203, "le Wis" is interpreted as "the wise"; in 1296, "le Reed" is interpreted as "the red".  Another common subtype was a metonymic byname, which identified a person's occupation by naming a tool used or a product produced or sold.  Thus, in 1296, "Thomas Mayle" could have been a maker of mail armor and in 1303, "Geoffrey wythe Hameres", viz., "with the hammers", was likely a maker or user of hammers (rather than a man who owned some notable hammers).  In addition, more complex characteristic bynames existed, although usage was generally less frequent than simpler forms.  Some English examples are found, such as, in 1159, "Shirloc", literally "'bright-lock", i.e., "fair-haired" ("Sherlock" in modern usage); in 1205, "de fer", or "iron-arm'; in 1209, "Yrento", literally "iron-toe"; in 1212, "le Gode" or "the good"; in 1219, "Cunteles" or "cunt-less" and "le Gidye" or "the mad" (in modern English, "giddy" has a much weakened sense); in 1249, "le Lechur" or "the lecher"; in 1251, "le Cruel", obviously "the cruel"; in 1275, "Wytelas" or "witless" and "le Wilfulle", obviously "the willful"; in 1286, "Foulmouth", no interpretation necessary; in 1290-92, "le Long", obviously "the long"; in 1293, "le Oneyede" or "the one-eyed"; in 1297, "Wythe Berd" or "with the beard" (actually misspelled as "Wychthe Berd"); in 1327, "Wysheued" or 'wise-head' and "Swetemouth" or "sweet mouth"; in 1375, "Notegood" or "not good"; in 1379, "Smalbyhind" or "small behind"'.  Obviously, as indicated above, some of these nicknames are complementary and some are non-complementary.  Other English characteristic bynames refer to articles of clothing, e.g., in 1297, "Wythemantel" or "with the mantle"; in 1307, "Wytebelt" or "white belt"; in 1312, "Scortmantil" or "short mantle".  Some cases are indistinct and not readily classifiable, such as in 1225,  "Gilbert Hodde" or "hood", which may be an occupational byname denoting Gilbert as a "maker of hoods", rather than just a characteristic byname, "Gilbert with the hood".  Likewise, in 1297, "Robert Rotenheryng" or "rotten herring", was likely a seller of fish.  Other nicknames are inconstent with preceding subtypes, such as in 1327, a man denoted as "le Wyfles", presumably had no wife.  Likewise, nicknames could apparently commemorate a particular event, such as in 1301, "Falinthewol" interpreted as "fall in the well", although this particular byname is found more frequently than might be expected, which suggests that it derives from some popular medival expression (now fallen out of use), perhaps, denoting a dreamer or a very clumsy person.  Likewise, in 1242, the rather cryptic byname "Latethewaterga" or "let the water go'', is ambiguous and allows many proper and profane interpretations, such as "the urinator or pisser" as well as many others.  In general, these complex, cryptic nicknames are quite rare and even absent in some European languages.  Of particular significance are English nicknames such as "Shakespeare" obviously "shakes spear", or in 1219, "Makepais" or "make peace"; in 1227 "Hakkeches" or "hack cheese" denoting a cheesemonger; or pejorative "petty criminal" subtypes such as in 1227, "Mangeharneis" interpreted as "eat harness", i.e., "eat armor"; in 1246, "John Fillecunt" or "fill the cunt" (who might have had difficulty with "Bele Wydecunthe" or "wide cunt" in 1327!); in 1269, "Brekebac" or "break back"; in 1275, "cuttepurs" or "cut purse" (at present, cutpurse is broadly equivalent to a purse snatcher) and "Badinteheved" or "bad in the head"; in 1279 "Aydrunken" or "always drunk; in 1327, "Strokelady" or "stroke lady" and "Brekaldoun" or "break all down" and 'Gobytheweye" or "go by the way"; in 1334, 'Brekelaunce" or "break lance".  Other types of phrases, sometimes even more elaborate, also appeared as nicknames; hence in 1246 one finds, "Fayrarmful" or "fair armful'; in 1282, "Drink al up", obviously meaning "drink all up'; in 1301, "Haldebytheheved" interpreted as "hold by the head" and "Fayrandgode" interpreted as "fair (handsome) and good"; in 1302, "Potfulofale" or "pot full of ale"; in 1309, "Singalday" or "sing all day"; and in 1211, the mysterious "Elias Overandover" or "over and over".  Concomitantly, in 1301 one finds the apparently married couple, "Henry Lytilprud" or "little worth" and his wife "Hawisia Crist a pes" or "Christ have peace', which probably records a favorite expression of this apparently long-suffering woman.
     In England, hereditary surnames were adopted in the thirteenth and fourteenth centuries, initially by the nobility, but eventually by everyone.  By 1400, many if not most English and some Scottish people used surnames, but many Scottish and Welsh people did not adopt surnames until the seventeenth century, or even later.  As might be expected, surnames generally derive from medieval bynames, which are conferred through sequential generations, rather than on just a single individual as originally.  By the sixteenth century, King Henry VIII decreed that marital births be recorded with the surname of the father.  In England and the later British Empire, there is a venerable tradition that a woman should change her surname upon marriage from her birth surname to her husband's surname.  Concomitantly, the first known instance in the United States of a woman insisting on retaining her birth name after marriage was that of Lucy Stone in 1855.  Subsequently, this practice has generally increased and became more popular in the 1960's, 1970's, and 1980's, but still remains a minority practice (as do hyphenated surnames consisting of a combination of maternal and paternal surnames seperated by a hyphen, blended surnames in which elements of paternal and materal surnames are combined into a single new surname, etc.).  Moreover, in the 1990's there has been an apparent decline in the percentage of birth name retention among married women.  Accordingly, as of the twenty-first century, more than 80% of American women still adopted her husband's surname after marriage.  Within this contest, modern European and American naming practices have evidently recovered the Roman tria nomina in the modified form of a first name, middle name, and surname.  Of course, the analogy is not exact since first and middle names really serve as duplicate prenomina and no analog of the cognomen is generally included.  (Of course the nomen and surname remain essentially analogous.)  For completeness, it may be observed that there is one European country, viz., Iceland, in which its citizens have never officially adopted surnames, but retain patronymic naming such that if a man named Karl has a daughter named Anna and a son named Magnús, their full names conventionally will be "Anna Karlsdóttir", viz., Karl's daughter, and "Magnús Karlsson", viz., Karl's son.  The feminine name is not changed upon marriage.

Return to Index