Sacred Texts  Legends & Sagas  Celtic  Index  Previous  Next 

p. 120

CHAPTER VIII.

THE CELTIC DIALECTS AND THE PROBABLE CHARACTER OF THE PICTISH LANGUAGE.

THERE is a fallacy which lurks in many of the arguments regarding the ethnological character of the old Celtic nations, based upon the modern languages. In arguing from the modern languages, it is always assumed that the language of each branch of the old Celtic race must be represented by one or other of the modern Celtic dialects. This fallacy pervades the writings of almost all of our ethnological writers, who argue as if, when a classical writer states that a difference existed between the language of two divisions of the old Celtic people, and when there is reason to suppose that the language of the one resembled the Welsh, then it must of necessity follow that the language of the other was the Gaelic. But this by no follows; nor is it at all self-evident that these modern Celtic languages represent all the ancient dialects. On the contrary, analogy and experience would lead us to a different conclusion. The ruder a language is, the more multiplied are its dialects; and the great medium for reducing their number is its cultivation. Before the introduction of writing, the means of such cultivation were to a great extent wanting. The Christian church was the great

p. 121

civiliser; and it was through its agency that these dialects received their cultivation, and one of their forms raised to the position of a written language. In the ante-Christian period of the Celtic language, the diversity of dialects must have been very great, and there may be many which have no direct representative among the modern languages. There may be many lost dialects on the Continent; and one such certainly existed, as we have seen in our own island, which has long ago disappeared--viz. the Pictish.

There run, however, through the whole of the modern Celtic, languages two great distinctive dialectic differences, which lie deep in the very groundwork of the language, and must have existed before their entrance into Great Britain, if not before their entrance into Europe. These differences separate these languages into two classes, each consisting of three of the spoken tongues. The one class, which we shall call the Cymric, consists of the Breton, the Welsh, and the Cornish;, the other, which we shall call the Gaelic, consists of the Irish, the Manx, and the Scotch Gaelic. The three Gaelic dialects are much more closely allied to each other than the three Cymric dialects; but each of the dialects composing the one class possesses in common those great distinctive differences which separate them from the three dialects composing the other class.

But while this great diversity exists, there are also analogies so close, vital, and fundamental, as to leave no doubt that they are all children of one common parent. Their vocabulary is, to a great extent, closely

p. 122

allied. A distinguished Welsh scholar of the present day estimates that two-thirds of the vocabulary of the six dialects are substantially the same; and I believe this conclusion to be correct. A number of the primitive adjectives expressing the simplest conceptions are the same. It is a peculiarity of both classes that the irregular forms bear a smaller proportion to the regular forms than is usual; but these irregular forms, which are, in fact, the deposit of an older stage of the language, bear a very remarkable analogy to each other.

The great and leading peculiarity in both classes of the Celtic languages, however, is the mutation of initial consonants; and while these initial mutations exist in each class, and are governed by the same laws, and thus afford additional evidence of their common origin, they at the same time present us with a means of discriminating between the different dialects, and distinguishing their mutual position as such, quite as effectual as Grimm's law has been among the German dialects. The consonants most readily affected by initial mutation are the mute consonants; and the following tables will show what the initial mutations in Welsh and Irish are:--

TABLE I.--INITIAL MUTATION OF MUTE, CONSONANTS.

p. 123

But while these consonants thus undergo a change according to fixed laws within the limits of the language itself, there is also a similar interchange of sounds between the different spoken languages; and it is obvious that if the changes which the same words undergo in different dialects follow regular laws, the phonetic laws of these languages are of the utmost importance in discriminating their dialectic differences. The phonetic law which governs the relations of Welsh and Gaelic, so far as regards the mute consonants, is this:--Each mute consonant in Welsh has two changes in Gaelic, either into its own middle sound, or into another consonant of the same character, but of a different organ. Thus the labial p passes into its middle sound b, as in

or into the guttural c, as in

[paragraph continues] This latter change is deeply rooted in Welsh and Gaelic, and enters into the very life of the language, of which we have two very remarkable instances. The word Pascha, for Easter, can only have entered these languages after the establishment of the Christian church, when the languages, under the influence of its teaching, were passing into the fixed form of a written and cultivated speech; but while in Welsh it becomes

p. 124

pasg, in Gaelic, under the operation of this law, it becomes casg. On the other hand, St. Ciaran, an Irish saint, and the founder of Clonmacnois, passed over, in the sixth century, into Cornwall, and had no sooner put his foot on Cymric ground than he became St. Pieran.

In the next class of the mutes the converse takes place, for the Welsh guttural g either disappears or passes into the dental d, as in

[paragraph continues] There is here, however, a slight deviation from the general rule: g in Welsh is usually combined with w, and is in this combination the Welsh digamma; but instead of passing into w, according to the law, it becomes in Gaelic f; that is, the guttural in Welsh passes into an aspirated labial in Gaelic, as in

This is sufficient to illustrate the law of this double change; but it is rather remarkable that while the one change is into a different character of the same letter, and in strict accordance with the phonetic change within the language itself, the other change is from a letter of one organ to that of another, as from labial to guttural, and guttural to dental. The operating cause

p. 125

of this rather startling change is to be found within in the laws which govern the sounds of the whole languages of this class, and in consequence of which the same phenomenon presents itself in other members of the Indo-European family.

There are two influences at work at all languages, antagonistic and mutually destructive of each other--the etymologic and the phonetic. The one governs the formation of a language, the other aids in its disorganisation. The etymologic influence has reference to meaning only, and brings together sounds which do not harmonise. These are immediately assailed by the phonetic influence, and modified till they are brought to a more simple and harmonious sound. History knows nothing of the formation of languages, and the phonetic influence is at work, and language in a process of decay, before the people which speak it have entered the historic period; but when these phonetic laws have become known, we are able to trace back the sounds, however impaired, to their original constituent elements. These contrasts, then, of labial and guttural, and guttural and dental, draw us back to a time when there were complex sounds which the human ear could not long tolerate, and which, by the modification of one or other element, passed over into the more simple sound, and in their divorce from each, other present this great contrast. There was probably a complex sound composed of a guttural and labial; k, or hard c, and v or p. By one member of the family the c will be softened to s, and then disappear; while the v will

p. 126

be hardened to p, and remain alone. In another, the hard c will remain, and the v be softened to u, and then disappear, leaving the c alone. An instance of this is the word for a "horse," which runs through most of the languages of the Indo-European family. The original term must have been acvas; in Sanscrit it becomes asvas; in Zend, aspas; in Greek, ippos; and in Gaulish or old Celtic, epo. In Latin the hard c is retained, and v modified, and it becomes equus; and in Gaelic, ech. The same process would seem to have been gone through within the Celtic languages, as the old inscriptions indicate that the old Celtic word for a "son" was maqvas. By one branch of the race the hard c was softened, and then dropped; while the v was hardened to p, producing the Welsh map (a son). By the other, the hard c was retained, but the v softened to u, in which form we have it as maqui, and finally dropped, leaving the Gaelic mac. The digamma, too, was originally a complex sound, which in Welsh is gw, and in Latin v, and in Gaelic f.

The consonantal changes between Welsh and Gaelic are, then, as follow:--

TABLE II.--PHONETIC LAWS BETWEEN WELSH AND GAELIC.

The vowel changes from Welsh to Gaelic are from w to o and y to e, which are likewise the masculine and feminine forms in Welsh, as--

p. 127

The vowel e becomes ea, as in pen (a head), ceann, and beann, G. Such being the relations between Gaelic, and Welsh, it must be obvious that they are of a nature to enable us to fix, from the form of the words, the relative position of almost any Celtic dialect to these two great types of the twofold division of the language; and the question at once arises, whether they may not enable us to determine the position of that one Celtic dialect in Great Britain--of which we have no direct living representative--viz. the Pictish. Of this language only five words have been handed directly down to us; but still, if these words are of such a kind as to exhibit some of the phonetic laws of the language, we are not without the means of determining this question. These five words are--

1. PEANFAHEL.--Bede, who wrote in the eighth century, says that the Roman Wall commenced about two miles west of the monastery of Abercorn, "in loco qui sermone Pictorum Peanfahel, lingua autem Anglorum Penneltun appellatur;" and Nennius adds that the wall was called "Britannico sermone Guaul," and extended "a Penguaul quæ villa Scotice Cenail, Anglice vero Peneltun dicitur." This gives us Penguaul as the British form, Peanfahel as the Pictish, and Cenail as the Scottish.

p. 128

2. UR.--One of the Pictish legends which had been added to the Historia Britonum, and has been preserved in the Irish Nennius, is expressly stated to have been taken from the books of the Picts, and has so important a bearing on this question that I insert it here entire:.--

"Of the origin of the Cruithneach here. Cruithne, son of Cing, son of Luctai, son of Partalan, son of Agnoin, son of Buain, son of Mais, son of Fathecht, son of Iafeth, son of Noe. He was the father of the Cruichneach, and reigned a hundred years. These are the seven sons of Cruithne--viz. Fib, Fidach, Fodla, Fortrend, warlike, Cait, Ce, Cirig--and they divided the land into seven divisions, as Columcille says:--

"Brude Pont, B. urpont, B. Leo, B. urleo, B. Gant, B. urgant, B. Gnith, B. urgnith, B. Fech, B. urfeich, B. Cal, B. urcal, B. Cint, B. urcint, B. Feth, B. urfeth, B. Ru, B. ero, B. Gart, B. urgart, B. Cind, B. urcind, B. Uip, B. uruip, B. Grith, B. urgrith, B. Muin, B. urnmin." 1

Thus ends this very curious fragment, which

p. 129

undoubtedly contains a number of Pictish vocables. I shall advert to these afterwards; at present I have to do with only one. It will be observed that the names of the thirty kings descended from Bruide Pont consist of only fifteen vocables, each name being repeated with the syllable ur prefixed. We have something exactly analogous to this in the old Welsh genealogies annexed to the Harleian MS. of Nennius, and written in the year 977. The ancestry of Cunedda Guledig is there thus given:--Cunedda, son of Patern, son of Tacit, son of Cein, son of Gwrcein, son of Doli, son of Gwrdoli, son of Duvn, son of Gwrduvn. This is evidently the same thing--guor, gur, or gwr, representing the Pictish ur. Again, one of the Pictish names is Urgest; and this name is repeated afterwards in the list of Pictish kings, where we twice have Ungust, son of Urgest; while the Irish Annals give the Irish equivalent as Aongus, son of Feargus--fear representing ur. We thus get the following forms:--Cymric, gwr; Pictish, ur; Gaelic, fear.

3. SCOLOFTH.--Reginald of Durham, in his Libellus de admirandis Beati Cuthberti Virtutibus--a work of the twelfth century--tells: of, a certain "Scolasticus Pictorum apud Cuthbrictiskchirch," or Kirkcudbright in Galloway; and says he was one of those "clerici qui in ecclesia illa commorantur qui, Pictorum lingua Scollofthes cognominantur." Scolasticus in Welsh is yscolheic; in Irish, sgolog.

4. CARTIT.--Cormac, in his old Irish Glossary, compiled in the ninth century, has--"Cartit, id est

p. 130

delg, id est belra cruithnech, id est delg for a curtar a choss;" that is; "cartit, a buckle, is a Pictish word. It is a buckle for putting on the foot." The Welsh equivalent is gwaell; the Irish is given by Cormac, dealg.

5. DUIPER.--In another of the Pictish fragments, which also formed part of the Pictish Chronicle, one of the mythic kings is thus given, "Gartnaidh Duiper." In the Chronicle of the Priory of St. Andrew, which contains a Scottish list of the same kings, the epithet is translated thus--"Gartnech dives," or rich. "Rich" in Welsh is goludog; in Irish, saoibher.

From these five words we gather the following phonetic changes. In the first we see the initial p in Cymric and Pictish passing over into c in Gaelic, the Cymric e passing into ea in Pictish and Gaelic, and the Cymric gu passing into f in Pictish, and neutralised by aspiration in Gaelic. In the second, gwr becomes ur in Pictish, fear in Gaelic. In the third we see the final guttural in Cymric and Gaelic softened to the dental in Pictish. The fourth is a peculiar word, but the Welsh and Irish equivalents furnish an example of g passing into d. In the fifth, the Pictish duiper and the Gaelic saoibher are the same word, showing d passing into s.

From these examples, Pictish appears to occupy a place between Cymric and Gaelic, leaning to the one in some of its phonetic laws, and to the other in others. Thus in the initial of the first word we have a Cymric form. The vowel-changes are Gaelic, and

p. 131

the initial of the second syllable also Gaelic; and on comparing the first two, words we see, that, while gw in Cymric ought, according to the general law; to pass into u in Gaelic--but in reality passes into f--the Pictish law combines both; and the Pictish canon is that gw in Cymric before a consonant becomes u in Pictish, and before a vowel becomes f in Pictish as in Gaelic.

The other words do not help us, at this stage of the inquiry; but we have another source of information in the proper names, of which we have in the lists of the Pictish kings the Pictish forms in the Irish Nennius and the Pictish Chronicle, and the Irish or Gaelic forms in the Chronicle of the Priory of St. Andrew and the Irish Annals, while the Welsh genealogies furnish Cymric equivalents. The phonetic laws which govern these, are equally available for our purpose. First, the Pictish law which changes gw into u before a consonant and f before a vowel, appears in the Pictish names Urgest, Uroid, and Fingaine; the Cymric equivalents of which are Gwrgust, Gwriad, and Gwyngenau; and the Gaelic, Feargus, Ferat, and Fingon. Then in the Pictish Drust, Deriloi, and Dalorgan, the Cymric equivalents of which are Grwst, Gwrtholi, and Galargan, we have the g passing into d, which is a Gaelic form. In the Pictish Domnall, the Cymric equivalent of which is Dwfnwall, we have the vowel-change of w into o, also a Gaelic form. The following table will show the result of this analysis:--

p. 132

TABLE III.--COMPARISON OF CYMRIC, PICTISH, AND GAELIC WORDS.

The Pictish tradition which I have given at length, besides yielding the word ur, furnishes us with a series of Pictish vocables. These are, first, the seven sons of Cruithne. They are said to have divided the land into seven portions, and to have given their names to them. We can identify some of them. "Fib" is plainly Fife, the old form of which was Fibh. "Fodla" is Atholl, the old form of which name was Athfodla. "Fortrenn" is the well-known name of the central district of the Pictish kingdom, which has now disappeared. "Cirig" or "Circin," as in the Pictish Chronicle, is the district of Girgin or Maghghirghin; now corrupted into Mearns, or Kincardineshire. "Caith" is Caithness, as in the old poem in the Irish Nennius,--

p. 133

that is, from Caithness to the Forth, the southern boundary of the Pictish kingdom. "Ce" and "Fidach" I cannot identify. But it will be observed, of these seven sons, the names of four begin with f, and the other three with c, obvious Gaelic forms; and I am inclined to think that they mark out a division of the Pictish race into two, of which one affected the guttural c, and the other the softer sound of the f.

Of the six names which follow, Aenbecan and Finecta are Gaelic forms; Guidid, Cymric; Gest, Urgest, and Brude, Pictish, as distinguished from either; and the untranslated epithets, Gadbre, Geis, and Gurid, are probably Pictish words.

The names of the thirty Brudes yield also fifteen Pictish monosyllables. These are, alphabetically, Cal, Cint, Cind, Fech, Feth, Gant, Gart, Geis, Gnith, Grith, Leo, Muin, Pont, Ru, Uip; and here also the prevalence of the gutturals, c, g, and the soft f is apparent. Some of these monosyllables have a resemblance to the names of the old Irish letters which signify trees, as cal, the name for c, a hazel; feth seems the same as pet, the name for p; gart, like gort (ivy), the name for g; muin, the vine, is the name for m; and leo resembles luis, and ru, ruis, ash and alder, the names for l and r. In the same manner three of the names of the seven sons of Cruithne have a resemblance to three of the numerals; as fib, pump, five; ce, se, six; caith, saith, seven. These, however, may be casual resemblances.

The relation of the fifteen vocables to the proper

p. 134

names is more apparent. On analysing the proper names of the Cymri and the Gael we find that both are produced by the same process--viz. a certain number of monosyllables forms the first half of the name, and to these are affixed a certain number of endings, the combination of which forms the proper names. In Cymric the initial syllables are--Ael, Aer, Arth, Bed, Cad, Car, Col, Cyn, Dog, Dygvn, El, Eur, Gar, Gor, Gwen, Gwyn, Gwyd, Gwr, Id, Mael, Mor, Tal, Tud, Ty. The Irish initial syllables are--Aen, Ain, Air, Aid, Art, Cath, Con, Corb, Cu, Domh, Donn, Dubh, Dun, Each, Echt, Eoch, Er, For, Fian, Fin, Finn, Fedh, Fear, Fail, Flaith, Flann, Gorm, Ir, Laigh, Lear, Lugh, Maen, Muir, Ragh, Reacht, Ruadh, Rud, Saer, Tuath. It would be endless to enumerate the affixes; but the most common Cymric are--deyrn, varch, wyr, swys; as, Aelgyvarch, Cadvarch, Cynvarch, Aerdeyrn, Cyndeyrn, Arthwys, Cynwys, etc.; and in Irish, cal, or in oblique case, gal and gusa; as, Aengus, Artgal, Ardgal, Congus, Congal, Dungus, Dungal, Feargus, Feargal, and so forth. Now these fifteen Pictish vocables likewise enter into the Pictish names, as Gart in Gartnaidh, and Dergart and Geis in Urgest; Leo in Morleo, Muin in Muinait, Uip in Uipog, and so forth. On the whole, the Pictish vocables coincide more with the Irish than with the Cymric, as Cal with Gal, Geis with Gusa, and so forth.

Further, on comparing the initial forms in Irish and in Cymric, we see in Cymric no words beginning with f, while in Irish there are nine; so that the vocables in

p. 135

[paragraph continues] Pictish with initial f are Gaelic. On the other hand, six vocables begin with g in Cymric, and only one in Irish; so that here the Pictish draws to the Cymric, and stands between the two with a greater leaning to the Gaelic.

The same fallacy which pervades the ethnological deductions regarding the Gauls also affects this Pictish question. It has been too much narrowed by the assumption that, if it is shewn to be a Celtic dialect, it must of necessity be absolutely identic in all its features either with Welsh or with Gaelic. But this necessity does not really exist; and the result I come to is, that it is not Welsh, neither is it Gaelic; but it is a Gaelic dialect partaking largely of Welsh forms.

It has always appeared to me that we can trace in the Celtic languages a twofold subordinate dialectic difference lying side by side, which is very analogous to some of the differences between high and low German. I do not mean to say that the differences between those subordinate Celtic dialects are absolutely parallel to those, between high and low German; but merely that they are of a nature which renders this nomenclature not inapplicable, while it affords a convenient term of distinction. A leading distinction between the high and low German is the preference of the latter for the sharp sounds, p, t, and k, instead of f or pf, s or z and ch; and the instance most familiar to us is the substitution of t for s, as wasser in high German becomes water in low, and water in English; dasz in high German is dat in low, and that in English.

p. 136

Now, a similar distinction is, in one point of view, observable among the three dialects of the Cymric. Of these dialects, the Cornish and Breton are much nearer to each other than either is to the Welsh. It is, in fact, a mistake to suppose, as is frequently asserted, that a Welshman and a Breton can understand each other. One of our best Welsh scholars, Mr. Price, who visited Bretagne, remarks: "Notwithstanding the many assertions that have been made respecting the natives of Wales and Brittany being mutually intelligible through the medium of their respective languages, I do not hesitate to say that the thing is utterly impossible. Single words in either language will frequently be found to have corresponding terms of a similar sound in the other, and occasionally a short sentence deliberately pronounced may be partially intelligible; but as to holding a conversation, that is totally out of the question." Cornish and Breton are much more nearly allied. Now, it is remarkable that in many cases d, dd, and t, in Welsh, pass into s in Cornish and z in Breton, as in

which is exactly analogous to one of the leading differences between high and low German; and Welsh, like the latter, shows a great preference for the dentals and its aspirates. I am therefore inclined to introduce the same nomenclature among the Celtic languages,

p. 137

and to call Welsh "low Cymric," Cornish and Breton "high Cymric" dialects.

The three dialects which compose the Gaelic class are much more nearly allied to each other than even Cornish and Armoric, and may be held to represent the old Scottish. On the same analogy they all belong to a high Gaelic dialect. There are, to be found, however, among the synonyms in the Gaelic dialects, low Gaelic forms accompanying high Gaelic forms, as in

which seems to indicate that a low Gaelic dialect has been incorporated or become blended with it.

The Pictish language appears to have approached more nearly to the old Scottish than even Breton to Welsh, according to Mr. Price's view; for Adomnan, who, in the seventh century, wrote the Life of St. Columba, the Scottish missionary to the Picts, describes St. Columba, the Scot, as conversing freely with the Picts, from the king to the plebeian, without difficulty; but when he preached to them the Word of God, he was obliged to make use of an interpreter: that is, he could make himself understood in conversing, but not in preaching; and, conversely, a Pict understood what he said in Scottish, but could not follow a Scottish sermon. This is a point, in fact, as to which there exists much misapprehension; and we are apt. to forget how

p. 138

very small a difference even in pronunciation will interpose an obstacle to mutual intelligence. Even in Breton and Cornish, the two Cymric dialects which most nearly approach each other, Norris, the highest Cornish authority, says, "In spite of statements to the contrary, the writer is of opinion that a Breton, within the historical existence of the two dialects could not have understood a Cornishman speaking at any length, or on any but the most trivial subjects;" and between Irish and Scotch Gaelic it would not require very much additional divergence to prevent the one from understanding the other.

Such being probably the mutual position of Pictish and Scottish, the few words we are able to compare show the difference between them to have been of the same character as between the high and low dialects; for we find saoibher (rich) in Irish represented by duiper in Pictish; and in proper names, Sarran by Taran, showing s in the one represented by d and t in the other; while the words sgolofth, cartit, and the proper names, Bargoit, Wroid, Wid, show the preference of the Pictish for dental in place of guttural terminations. I consider, therefore, that Pictish was a low Gaelic dialect; and, following out the analogy, the result I come to is, that Cymric and Gaelic had each a high and a low variety; that Cornish and Breton were high Cymric dialects, Welsh low Cymric; that old Scottish, spoken by the Scotti, now represented by Irish, Scotch Gaelic, and Manx, was the high Gaelic dialect, and Pictish the low Gaelic dialect.

p. 139

This analogy is confirmed by the legendary origins of these different races, in which, under the form of a mythic migration, the traces of a rude and primitive ethnology often lie hid. The tendencies which produce the high and low German are, as we have remarked, associated with the character of the country peopled by them. The low German forms are connected with the level and marshy plains which border on the German Ocean, the high German with the more mountainous region of the south of Germany; but the same characteristics mark the mythic. migrations of the Celtic races which peopled Britain. In the Welsh traditions, the Cymry, which are represented by the Welsh or low Cymric people, are said to have crossed the German Ocean from the north of Germany; the Lloegrys, represented by the Cornish or high Cymric, are brought from the South. In the old Irish traditions, the different races said to have peopled Ireland fall into two classes: the one is said to have penetrated through Europe by the Rhiphaean Mountains to the Baltic, and to have crossed the German Ocean; and the other is brought by the Mediterranean and the south of Europe. 1 The former alone are said to have made settlements in Scotland; and Bode, in giving the tradition of the origin of the Picts, brings them likewise from the north of Germany across the German Ocean. This population which preceded the German races was, in fact, the race of the Celts, who seem to have been driven westward by the

p. 140

pressure of the Teutonic movement; and, like the German, to have shown a twofold minor difference, produced by the same physical influence, which is known by the names of "high" and "low" German.

The platform occupied by the Pictish people was not confined to Scotland alone, for they certainly extended over part of the north of Ireland, and formed, in all probability, an earlier population of the north half of Ireland, which became subjugated by the Scots. On the other hand, the Scots at an early period occupied the district of Argyll. In the north of Ireland and the west of Scotland the Picts must, at an early period, have become blended with the Scots, and their form of the Gaelic assimilated to the Scottish. In Scotland, south of the Tay, where they occupied the districts from the Tay to the Forth, the region of Manau or Manann, and Galloway, they came in contact with the Cymric people, and the one being a low Gaelic dialect, and the other a low Cymric dialect, their forms must have so far resembled each other as to lead to an admixture presenting that mixed language of low Gaelic with Cymric forms, known to Bode as the Pictish language.

Footnotes

128:1 Chron. Picts and Scots, p. 24.

139:1 The one class consists of the Nemedians and the Tuatha de Danaan; the other of Partholan and his colony, the Firbolg and the Milesians.