Surnames as a Science

It happens, perhaps yet more frequently, that a German name, which cannot be explained by anything within the range of Teutonic dialects, may find a sufficient etymon from the Celtic. That is to suppose that a word originally common to the Teutonic and the Celtic, has dropped out of the former, and been retained only in the latter. Thus there is a word arg, arch, found in many Teutonic names, and from which we have several names, as Archbold, Archbutt, Archard, Argent, Argument, for which the meaning that can be derived from the German seems very inadequate, but for which the Irish arg, hero or champion, seems to offer as good a meaning as could be desired. So also all, from which, as elsewhere shown, there are a number of names, in its Teutonic sense of omnis, does not seem to give by any means so satisfactory a result as in its Celtic sense of "great" or, "illustrious." Many other instances might be adduced on both sides to show the way in which a word has dropped out of the one language and been retained in the other.

Before passing from this part of the subject, I may be allowed to adduce an illustration – a striking one I think, albeit that the name in this case is not that of a man but of a dog – of the way in which a name may be retained in familiar use, though the word from which it is derived has perished out of the language, though the language itself has passed out of use among us for more than a thousand years, and though the word itself is only used in a sort of poetical or sentimental sense. Who has not heard, in verse or in prose, of the "poor dog Tray"? And yet who ever heard, excepting in books, of a dog being called Tray, a word which conveys no meaning whatever to an English ear? What then is the origin, and what is the meaning, of the name? It is, I venture to think, the ancient British name for a dog, which is not to be found in any living dialect of the Celtic, and which is only revealed to us in a casual line of a Roman poet: —

Non sibi, sed domino, venatur vertragus acer,

Illæsum leporem qui tibi dente feret.

Martial.

The British vertrag must have been something of the nature of a greyhound, though, from the description of his bringing back the game unmangled to his master, perhaps capable of a higher training than the greyhound generally attains to. Now the ver in vertrag is in the Celtic tongues an intensitive, and as prefixed to a word, gives the sense of preeminence. The ancient British word for a dog in general must have been trag, a word of which we find a trace in the Irish traig, foot, allied, no doubt, to Gothic thragjan, Greek τρεχειν, Sanscrit trag, to run. The ancient British name then for a dog, trag signified the "runner," and with the intensitive prefix ver, as in vertrag, the "swift runner."6 And trag is, I take it, the word from which, g as usual in English becoming y, is formed our word Tray.

It may be of interest, in connection with the antiquity of our names, to take a few of the oldest Teutonic names of which history gives us a record, and endeavour to show the relationship which they bear to our existing surnames. It will be seen that not only have we the representatives of these ancient names, but also in certain cases names which represent a still more ancient form of the word.

And first let us take the name, dating back to the first century of our era, of the old German hero Arminius, brought before us with such magnanimous fairness by Tacitus. The old idea, let me observe, that Armin is properly herman, leader or warrior, has long been given up by the Germans. The name, of which the most correct form is considered to be Irmin, is formed from one single word of which the root is irm, and the meaning of which is, as Grimm observes, entirely obscure. We have then as English surnames Armine, Ermine, and Harmony, the last, no doubt, a slight corruption, though, as far as the prefix of h is concerned, it is as old as Anglo-Saxon times, for we find "Harmines den," Harmine's valley, in a charter quoted by Kemble. Then we have compounded with gar, spear, and corresponding with an O.G. Irminger —Arminger, Irminger,7 and again as a corruption, Iremonger. And, compounded with hari, warrior, and corresponding with an O.G. Irminhar, we have Arminer. And, as a Christian name of women, one at least of our old families still retains the ancient name Ermentrude, the ending trude, as found also in Gertrude, being perhaps from the name Thrud, of one of the Valkyrjur, or battle-maidens of Odin. The French also, among the many names derived from their Frankish ancestors, have Armingaud, Armandet, and Ermingcard, corresponding with the ancient names Irmingaud, Irmindeot, and Irmingard. And Irminger, as I write, comes before me in the daily papers as the name of a Danish admiral. But Irmin is not the oldest form of the name, – "the older and the simple form," observes Foerstemann, "runs in the form Irm or Irim," and with this also we can claim connection in our family names. For we have the simple form as Arms and Harme; and as compounds we have Armiger, corresponding with an O.G. Ermgar; Armour, with an O.G. Ermhar; and Armgold, with an O.G. Ermegild. Lastly, I may observe that both Irm and Irmin are found also by Stark as ancient Celtic names. And certainly there is no stem more likely than this, of the origin of which all trace is lost in the darkness of the past, to be one that is older than the Arian separation.

The name Sigimar, of the brother of Arminius, I have already shown that we have, not only in its own form as Seymore, but also in its High German form as Sycamore, the Anglo-Saxon names from which they may be taken to be more immediately derived being also found in the chapter on place-names. And I have also shown that we have the name Cariovalda (or Harwald) of a prince of the Batavi, of the first century, in our Harold.

There was another old hero of the German race, not so fortunate as Arminius in finding an historian in a generous foe, whose name only comes before us in a line of Horace: —

Occidit Daci Cotisonis agmen

Cotiso must have been a leader of some High German tribe, perhaps somewhere on the Upper Danube, and he must have made a gallant stand against the Roman arms, inasmuch as his final overthrow is deemed by the Roman poet a worthy subject on which to congratulate his imperial patron. Cotiso is a High German form of another name, Godiso or Godizo, elsewhere referred to, and hence may be represented, I venture to think, in our names Godsoe and Goddiss, while Cotiso itself may be represented in our Cottiss, the ancient vowel-ending being in our names, as I shall show in the next chapter, sometimes dropped and sometimes retained.

Another name which goes back to the first century of our era is Arpus, that of a prince of the Catti in Tacitus. The Eorpingas, descendants or followers of Eorpa, were among the original settlers, and seem to have confined themselves to Norfolk, where alone we have any traces of them. The name may perhaps be referred to Anglo-Saxon eorp, wolf, though other derivations have also been proposed. We have the name at present as Earp (the name of a member of the House of Commons), and also as Harp. Upon this stem is formed the name Arbogastes (gast, guest) of a Frankish general under the Emperor Gratian in the fourth century; and Arbogast is still a family name among the French.

Lastly, let us take the name of the German king, Ariovistus, brought before us by Cæsar. The proper form of this name, there seems little doubt, is Arefastus, as found in some other O.G. names. There was also an Arfast, bishop of East Anglia, in the time of William the Conqueror. And Arfast is a present name among the Frisians, according to Outzen, who compares it – rightly, as it seems to me – with the old name Ariovistus. The corresponding name Arinfast (aro, arin, eagle) was also in ancient use among the Danes. It seems to me that our name Harvest may easily be a corruption of Arfast; it has generally no doubt been derived from a man's having been born at such a season, but I distrust, as a general rule, as elsewhere stated, derivations of this kind.

In connection with the subject of the antiquity of Teutonic names generally, and of English names as derived from them, I shall have, in a subsequent chapter, to refer to the names of original settlers in England as deduced by Kemble from ancient charters, and compare them with names of a similar kind found in Germany. The coincidence that will be found in these names at that early period, from England and Friesland in the north to Bavaria in the south, will, I think, be a very strong argument to show that these names could not have originated within the Teutonic area itself, and so dispersed themselves over it in its length and breadth, but that they must have been brought with them by the Teutonic invaders from their earlier homes.

CHAPTER II.

CLUE TO SOME OF THE ANCIENT FORMS REPRESENTED IN ENGLISH NAMES

So long as our surnames are treated as if each name were something standing apart by itself, very little progress can be made in their elucidation; it is by collation and comparison that, in this as in any other science, definite results are to be obtained. And a moderate amount of attention to the forms in which these names appear, and to the various endings prevalent among them, will enable many names, otherwise unrecognisable, to be brought within the pale of classification and of possible explanation. I am of course referring to that portion of our surnames – a much larger one according to my judgment than is generally acknowledged – which dates back to Anglo-Saxon times, and so forms a part of the general system by which Teutonic names are governed.

I shall have, in the course of this work, frequently to refer to the Teutonic system, and to names which do, or do not, according to my judgment, enter into it. And I will therefore, before going further, endeavour to explain what I mean by the Teutonic system. There is, then, a class of words which, at a time of remote antiquity, have been adopted as stems upon which, in some cases by a sort of phonetic accretion, in some cases by the addition of a diminutive ending, in some cases by forming a patronymic, in some cases by taking in another word as a compound, a number of other names have been formed. Thus, when we find such a group of names as Dill, Dilly, Dillow, Dillen, Dilling, Dilke, Dilwyn, or as Budd, Budden, Buddle, Budding, Buddrich, Budmore, we may take it that these are all ancient names, of which Dill and Budd are respectively the stems. And whenever we find a group of names with endings such as it is my object in the present chapter to explain, and in compounds such as will be dealt with in a succeeding chapter, we shall be warranted in assuming the antiquity of the group.

The endings in a, ay, ah, ey, ie, o, oe, ow

And in the first place, let us take the endings in a, i, and o, of which the above are nothing more than arbitrary variations of spelling. Now ancient Teutonic names formed of one single word had commonly, though not invariably (and the same thing applies also to ancient Celtic names), a vowel-ending in a, i, or o; this ending is in our names sometimes dropped and at other times retained. (It is to be observed, however, that even in Anglo-Saxon times it is not an unfrequent thing to find the same name variously with and without a vowel-ending, of which some instances may be noted in Chapter V.) Thus we have Abbe, Abba, and Abbey, we have Bell, Belly, and Bellow, we have Earl and Early, we have Dand, Dandy, and Dando, we have Brand and Brandy, we have Todd and Toddy, we have Dane and Dana, we have Marr, Marry, and Marrow. These are all ancient names, variously with and without the vowel-ending, and it will be readily seen how apt the addition is to disguise the name, and to give it the appearance of something else.

The question now to consider is – What is the value and meaning of this vowel-ending, which was only given to simple names and never to compounds? It might be, in some cases, used simply as a sort of euphonic rounding-off of a name which might seem meagre and insignificant without something of the sort. We ourselves appear to use s in the same manner in the case of some very short names, such as Wills and Epps, in which the final s may perform the same service that was rendered by the vowel-ending. But there is also another principle which I think obtains, and which, indeed, may be the guiding principle in such cases. In Anglo-Saxon (and the same principle applied to other Teutonic dialects), the addition of a to a word implied connection with it. Thus, from scip, a ship, is formed scipa, one connected with a ship, a sailor. Now, going back to the remote origin of names, there were many cases in which a man took a name from an abstraction, such as war, peace, glory, victory, or from a weapon, as the sword or the spear, and it is obvious that in such cases he required something to connect his name with it, and this is, as it seems to me, what was effected by the ending in question. And the principle is still a living one among us, and we form names daily in accordance with it, though we no longer use the ending in a, which has been superseded by that in i.8 A connection with anything whatever is expressed by this ending, as when a stupid person is called "Duncey," one with a remarkable nose "Nosey," or one with a halting gait "Stumpy." The French seem to have retained their old ending, and, when they form names of this sort, to do it with the ending in o (eau) which appears to be in accordance with the genius of their language, as that in i (ey) is with that of ours.

Of these three endings, that in a is the one which was in use among the Goths, in such names as Cniva, Totila, Ulfila. And the same was also the case among the Saxons, a branch of the same Low German stock, in such names as Anna, Ella, Penda, Dodda. The ending in i was also common among the Old Saxons, and, if we may judge by the Liber Vitæ of Durham (which might naturally be supposed to contain a large proportion of Northern names), was also prevalent in the ancient Northumbria. We have in that record the names Alli, Arni, Bynni, Betti, Cyni, Diori, Elsi, Paelli, Tidi, Tilli, Terri, all of which are found in our present names Alley, Arney, Binney, Betty, Kinney, Deary, Elsey, Paley, Tidy, Tilley, Terry. The ending in o was that which was in favour among the Franks and the High Germans generally, the oldest instance on record being probably that of Cotiso, p. 20. This is the usual ending in French names (so far as they are of Old Frankish origin, and come under this head), the form being generally eau, as in Baudeau, Godeau, Fredeau, representing the ancient names Baldo, Godo, Fredo. Hence our names ending in o may be taken to be, to some extent, names of Old Frankish origin come to us through the Normans. But the number of such names is larger than could reasonably be accounted for in such a way, and in point of fact, we meet occasionally with such names at a much earlier period. The Frisians certainly seem to have had names in this form, and it is a question whether such names may not be partly due to them. It must be observed, then, that names with these three various endings represent the stem just the same as those that are without it.

The ending in an, en, in, or on

This ending runs through the whole range of Teutonic names, and is common in English surnames. Hence we have Doran, Lingen, Bolden, Hannen, Farren, the names on which they are formed being represented in Dore, Ling, Bold, Hann, Farre. As to the value and meaning of this ending, we have nothing more to guide us than its parallel use in the languages most nearly concerned, where it is what may be called formative. That is to say, it is a form of speech which is used to form the endings of words, not adding anything to the meaning, but forming a kind of euphonic rounding-off of the word. Thus from A.S. wearda is formed warden, from geard (inclosure) is formed garden, from Brytta is formed Briton, from mægd, maid, is formed maiden. Cf. also the old word ratten for rat, still used in provincial speech. In many cases in Teutonic names we have words thus formed, and also the simpler forms on which they have been founded, e. g. we have bero, bear, and also berin, we have aro, eagle, and also arin (=A.S. earn), both forming the stems on which a number of other names have been built. I take the ending in en, then, to be most probably a kind of phonetic accretion, adding nothing to the sense, but sometimes representing a secondary word, and starting a stem on its own account.

The ending in ing

This is the Anglo-Saxon and ancient German patronymic, as in Browning, "son of Brown," Dunning, "son of Dunn," Winning, "son of Winn." It must have been superseded during, or very soon after, Anglo-Saxon times, by the patronymic in son, inasmuch as no names of Scriptural origin appear to be formed with it. Hence we have such names as Bulling, Burning, Canning, Gambling, Halling, Harding, Hopping, Loving, Manning, Swearing, Telling, Walking, Willing, some of which have been popularly supposed to be from the present participle. All of the above except two, Swearing and Gambling, are found in the list of early Saxon settlers, and of these two (which are found in after Anglo-Saxon times) Swearing, which corresponds with an Old German Suaring, finds its stem in an Anglo-Saxon name Sweor, signifying important, honourable; and Gambling (properly Gamling) is the patronymic of an A.S. and O.N. name, Gamol, signifying "old," probably in the honorific sense of old descent. From this origin, I take it, are also our names Farthing and Shilling, the former from the stem fard, or farth, signifying "travel," found in several ancient names, and which I rather take to be the same as ford, found in the Fordingas among the early settlers. And Shilling, which corresponds with a present German Schilling, is probably the same as the Scilling in the "Traveller's Song," a supposed contraction of Scilding, from A.S. scyld, shield, in which case our name Shield would be the parent of Shilling. I have referred at the beginning of this book to the curious-looking name Winfarthing (quoted from Lower) as perhaps a corruption of an A.S. Winfrithing, though it is a case in which I do not feel much certainty, finding one or two other such names as Turnpenny, which may have been sobriquets.

The ending in el or il

This ending in Teutonic names may be taken, as a general rule, to be a diminutive, though in a few cases it may be more probably, like that in en, formative. Thus in the list of early A.S. settlers we have Bryd(ingas) and we have Brydl(ingas), representing the words bride and bridle. Now, as German writers have taken the word brid in ancient names to mean "bridle," comparing it with French bride, it would seem probable that, in the above A.S. name, Brydl is not a diminutive, but the extended word "bridle." However, as a general rule, it may be presumed to be a diminutive, and in such sense I take the following, premising that this, as well as all other diminutives, except kin, lin, and et, is subject to a vowel-ending just the same as simple forms. We have Bable, corresponding with an A.S. Babel, and an O.G. Babilo; Ansell and Anslow (Ansilo), corresponding with an O.G. Ansila; Mundell and Mundella, with a Gothic Mundila;9 Costall, Costello, and Costly, with an O.G. Costila. Costly is properly Costili, with the ending in i, as also Brightly is Brightili, and some other names with an adverbial look may be similarly explained.

The ending in ec or ic

This ending, with rare exceptions, may also be taken to be a diminutive. The oldest instance on record is stated by Stark as that of the Vandal general Stilicho in the fourth century, though, as found on Roman pottery (in the names Bassico and Bennicus), it may be still older. It seems rather singular that, though, according to Grimm, this ending was more particularly in favour among the Saxons, not a single instance of it occurs among the names of our early settlers, nor indeed any other form of diminutive except that in el, though the form in question is not uncommon in after Anglo-Saxon times. This diminutive is still in living use among us, at least in Scotland, where a "mile and a bittock" (little bit) has proved a snare to many a tourist. We have Willock, Wilkie, and Wilke, corresponding with an O.G. Willico, and an A.S. Uillech; Lovick and Lubbock, corresponding with O.G. Liuvicho; Jellicoe, corresponding with O.G. Geliko, Jeliko, and an A.S. Geleca, some of these examples being with, and some without, the vowel-ending.

The ending in lin

This ending, which is also a diminutive, is probably formed from that in el, by the addition of en. It is found in Foerstemann's list as early as the fifth century, but, as found on Roman pottery, must probably be still older. We have Bucklin, corresponding with a Buccellin, general of the Alemanni in the sixth century, and with a Buccellan on Roman pottery. Also Tomlin, corresponding with an O.G. Domlin; Applin, with an O.G. Abbilin; Franklin, with an O.G. Francolin; Papillon, with an O.G. Babolen, &c. This form of diminutive never takes a vowel-ending.

The ending in kin

This diminutive ending is formed from that in ec by the addition of en. It is the youngest-born of all, not being found, unless in rare cases, before the tenth century. And it is one that is still in living use both in England and in Germany, in the latter country more especially. We have Wilkin, corresponding with an O.G. Williken, and an O.N. Vilkinr; Godkin, with an O.G. Gotichin; Hipkin, with an O.G. Ibikin or Ipcin; and Hodgkin, with an A.S. Hogcin.

The ending in et

There is an ending in d or t in O.G. names, which may be taken, though perhaps not with anything like certainty, to have the force of a diminutive. Hence might be such a name as Ibbett, corresponding with O.G. names Ibed and Ibet, from an unexplained stem ib; also our names Huggett, Howitt, and Hewitt, corresponding with an Anglo-Saxon Hocget, and an O.G. Huetus, from the stem hog, hug, signifying study or thought. But some other endings are so liable to intermix, and particularly the common one had, war, that there is very seldom anything like certainty.

The ending in es or is

I take this ending also to be diminutive, and to be possibly akin to our ish, as in blue-ish, which, as signifying a "little blue," seems to have the force of a diminutive. Hence we have Riches, corresponding with an O.G. Richizo, and a present French Richez; and Willis, corresponding with an O.G. Willizo. Then we have Godsoe, corresponding with an O.G. Godizo, of which Cotiso, mentioned in Horace (p. 20), is a High German form; and Abbiss, corresponding with the name, Abissa, of the son of Hengest, from, as supposed, Gothic aba, man. And we have Prentiss, corresponding with an A.S. Prentsa (=Prentisa), respecting which I have elsewhere suggested that the name should be properly Pentsa. Another name which I take to be from this ending is Daisy. There is an A.S. Dægsa, which as Dagsi, with the alternative ending in i, would give us Daisy. We have another name, Gipsy, which I take to be from Gibb or Gipp (A.S. geban, to give) with this ending. This ending in is is naturally very apt to be corrupted into ish, and it is from this source, I take it, that we have such names as Radish, Reddish, Varnish, Burnish, and Parish, the two last of which we have also in their proper form as Burness, and Parez or Paris.