The Celtic languages are unusual in the Indo-European family for two reasons. They are, first, a good deal less varied than the family as a whole — the Celtic languages are, in linguistic terms, remarkably similar to one another, and the differences between, say, Welsh and Irish are far smaller than the differences between, say, English and Hindi. They are, second, a good deal less well documented than most other Indo-European families: Latin has a continuous written record from the seventh century BCE, Greek from the eighth, Sanskrit from the second millennium BCE, but the Celtic languages have almost no written record before the Christian era, and the continental Celtic languages have almost none at all. The Celtic-speaking world was, on the whole, a world without writing until the Romans and the Christian missionaries arrived, and the result is a written tradition that begins late and depends on the accidents of inscription, manuscript, and survival.
The standard introductions to the Celtic languages are David Stifter’s Old Irish (2018) and Ranko Matasović’s A Concise Historical Grammar of Old Irish (2021). For the comparative linguistics of the family as a whole, Kim McCone’s Towards a Relative Chronology of Ancient and Medieval Celtic Sound Change (Maynooth, 1996) and the essays collected in Martin J. Ball and Nicole Müller (eds.), The Celtic Languages (Routledge, 2nd ed. 2009), are the standard references. For Ogham, Damian McManus’s A Guide to Ogam (1991) and R. A. S. Macalister’s older Corpus Inscriptionum Insularum Celticarum (4 vols., 1945) are the indispensable works. For the medieval literary tradition, the most useful single introductions are Robin Flower, The Irish Tradition (1947, repr. 1994), Patrick K. Ford, The Mabinogi and Other Medieval Welsh Tales (1977), and Joseph Clancy, The Earliest Welsh Poetry (1971).
The Place of Celtic in the Indo-European Family
The Celtic languages are one of the branches of the Indo-European language family, the great grouping of related languages that includes most of the languages of Europe, the Iranian plateau, and the northern Indian subcontinent. Within Indo-European, the Celtic languages are most closely related to the Italic languages (Latin, Oscan, Umbrian, and their relatives), the Germanic languages, and the Balto-Slavic languages, with the Italo-Celtic hypothesis suggesting an even closer relationship between Celtic and Italic in the prehistoric period. The Italo-Celtic hypothesis is supported by a number of shared sound changes and morphological features (the -r passive, the superlative suffix -isamo-, the treatment of the Indo-European ē) that are not found elsewhere in Indo-European, and it is now generally accepted as a real if distant relationship.
The Celtic languages share a number of distinctive features that set them apart from other Indo-European branches. The most famous of these is the loss of the original Indo-European p sound, a sound change that distinguishes Celtic from most other Indo-European languages. The Latin word for father is pater, the Greek patēr, the Sanskrit pitár, but the Old Irish word is athair and the Welsh word is tad. The same change shows up across the whole Celtic family and is one of the defining features of the group. The Celtic languages also have a system of consonant mutations, in which the initial consonant of a word changes under certain grammatical and syntactic conditions, a feature that is shared with very few other languages. The verb system of the Celtic languages is also distinctive, with a strong tendency to use verbal nouns, a complex set of absolute and relative verb forms, and a number of unusual constructions that have been the subject of intense linguistic study. The so-called Welsh subordinate construction and the use of the copula in cleft sentences, for example, are features that have no clear parallel in any other Indo-European language.
The earliest evidence for the Celtic languages as a separate group comes from the late Bronze Age or early Iron Age, when the speakers of proto-Celtic began to develop the distinctive features that would set them apart from the other Indo-European languages. The exact date and place of this emergence, the relationship of proto-Celtic to the older Indo-European dialects, and the chronology of the spread of the Celtic languages into the western fringes of Europe are all matters of active scholarly debate, and the Celtic language family examines these questions in detail.
The Continental and Insular Branches
The Celtic languages are conventionally divided into two main branches: the continental Celtic languages, which were spoken across most of continental Europe in the Iron Age and the Roman period, and the insular Celtic languages, which are spoken in the British Isles and (through migration) in Brittany. The continental Celtic languages are entirely extinct as living languages, but they are known from a substantial body of inscriptions, place names, personal names, and loanwords in Latin and other languages. The insular Celtic languages, although reduced in number and in speakers, are still spoken today.
The continental Celtic languages include Gaulish, which was spoken across Gaul and parts of northern Italy in the Iron Age and the Roman period and is known from hundreds of inscriptions and from a small number of manuscript texts. Celtiberian, spoken in north-central Spain, is known from a substantial corpus of bronze inscriptions. Lepontic, the oldest attested Celtic language, is known from inscriptions of the sixth century BCE in northern Italy. Galatian, spoken in central Anatolia by the descendants of the Celtic migrants of the third century BCE, is known mainly from a small number of personal names and from a single short text. Other continental Celtic languages, including Noric, Pannonian, and Galatian, are known from limited evidence.
The insular Celtic languages are divided into two sub-branches. The Goidelic, or Q-Celtic, branch includes Irish, Scottish Gaelic, and Manx, all of which preserve a k (or q) sound in words where the Brythonic languages have a p. The Brythonic, or P-Celtic, branch includes Welsh, Cornish, Breton, and the extinct Cumbric of the old kingdom of Strathclyde, all of which share a common ancestor usually called Common Brythonic. The split between Goidelic and Brythonic is generally dated to the late Bronze Age or early Iron Age, and it reflects a separation between the Celtic-speaking communities of Ireland and those of Britain that was well established by the time of the Roman conquest. The most striking difference between the two branches is the treatment of the original kw sound, which becomes p in the Brythonic branch and k (or c) in the Goidelic branch, and which appears in the continental languages as p in some positions and kw in others. The kw/p split has been the subject of a long-running debate, and the answer to the question depends partly on the reconstruction of proto-Celtic and partly on the interpretation of the continental evidence.
Irish, the Oldest Insular Tradition
Irish, the oldest of the insular Celtic languages, has the longest continuous literary tradition of any Celtic language. The earliest Old Irish texts date from the seventh and eighth centuries CE, although the material they contain is often much older, and the manuscript tradition that preserves them includes works from the early medieval period through the modern era. The earliest Old Irish glosses, brief notes written in the margins of Latin religious manuscripts, give a vivid picture of the spoken language of the early medieval period and show that by the seventh century Irish had developed a sophisticated literary and grammatical tradition.
The medieval Irish literary tradition is one of the most remarkable in Europe. The great manuscripts of the eleventh and twelfth centuries, including the Book of the Dun Cow (Lebor na hUidre, c. 1106), the Book of Leinster (c. 1160), the Yellow Book of Lecan (c. 1390), and the Great Book of Lecan (c. 1418), preserve a vast body of prose and verse that includes the great mythological cycles, the heroic tales, the historical sagas, the legal texts, the saints’ lives, and the lyric poetry of the early and middle Irish period. The richness and variety of this material, much of which was composed in the oral tradition centuries before it was written down, is one of the great treasures of European medieval literature.
The early Irish poets and scholars who worked in the bardic schools developed a highly sophisticated system of metrics, grammar, and literary composition that became a model for the later bardic traditions of Wales and Scotland. The bardic schools trained students in the complex syllabic verse forms of the classical Irish tradition, in the composition of prose tales in the classical style, and in the mastery of the vast body of traditional lore that was the inheritance of every professional poet. The poets who emerged from this training enjoyed a privileged position in medieval Irish society: they could compose poems of praise that honoured a king, poems of satire that could ruin an enemy, and elegies that preserved the memory of the dead.
Welsh, Breton, and the Brythonic Tradition
The Brythonic Celtic languages, although not as well preserved in the early medieval period as Irish, produced their own rich literary tradition. The earliest Welsh texts, including the Old Welsh glosses of the eighth and ninth centuries, show that the Brythonic language of Wales was already well developed as a literary medium by the time of the Anglo-Saxon settlement. The great Mabinogion, a collection of prose tales preserved in the Red Book of Hergest (c. 1400) and the White Book of Rhydderch (c. 1350), draws on a much older oral tradition that probably goes back to the early medieval period.
The Mabinogion contains some of the great stories of medieval European literature, including the four branches of the Mabinogi proper — Pwyll, Branwen, Manawydan, and Math — the great Arthurian tales of Culhwch and Olwen and the Dream of Macsen Wledig, and the medieval romances of Owain, Peredur, and Geraint. The Welsh Triads, a collection of grouped lists of persons, places, and events preserved in manuscripts of the thirteenth and fourteenth centuries, preserve a great deal of older mythological and legendary material. The Welsh bardic tradition, particularly the tradition of the prydydd or professional poet that flourished from the twelfth through the fourteenth centuries, produced an extraordinary body of formal verse that combines praise of patrons, religious poetry, and elegy.
Breton, the Brythonic language spoken in Brittany, has a much more limited manuscript tradition than Welsh or Irish, but it preserves a number of medieval texts, including the Life of Saint Columba and a small number of religious and literary works. The Breton oral tradition, which continued to flourish into the modern period, preserves a great deal of older material, and the gwerz, a form of Breton ballad, preserves stories of heroes, saints, and historical figures that often have older parallels in the Welsh and Cornish traditions. The most famous product of the Breton tradition in the modern period is the work of Théodore Hersart de La Villemarqué, whose Barzaz Breiz (1839, 1845), although it has been criticized for editorial liberties, has been enormously influential in shaping the modern Breton cultural identity.
Cornish, the Brythonic language of Cornwall, has a limited medieval manuscript tradition, but it preserves a few important texts including the Ordinalia, a cycle of medieval mystery plays. Cornish declined through the early modern period and was effectively extinct by the late eighteenth century, although it has since been revived as a community language. Cumbric, the Brythonic language of the old kingdom of Strathclyde and the southwestern Scottish lowlands, is known mainly from place names, personal names, and a small number of glosses, and it appears to have been extinct by the twelfth century. The Cumbric evidence is particularly important for the reconstruction of Common Brythonic.
Scottish Gaelic and the Goidelic Tradition Outside Ireland
Scottish Gaelic, the Goidelic language of the Scottish Highlands, has a literary tradition that begins in the late medieval period, with a small number of texts preserved in manuscripts of the fifteenth and sixteenth centuries. The earliest major text is the Book of the Dean of Lismore, a manuscript compiled in the early sixteenth century that contains a substantial collection of Gaelic poetry, much of it composed centuries earlier. The Scottish Gaelic tradition draws heavily on the older Irish tradition, and many of the tales, poems, and formulas of the medieval Irish repertoire continued to circulate in Scotland long after the two traditions had begun to diverge linguistically.
The Highland bardic tradition of the seventeenth and eighteenth centuries produced a remarkable body of formal verse, including the dan direach (strict syllabic verse) and the ornaid (ornamental verse). The poets of the Highland clearances and the Highland regiments preserved and adapted the older bardic forms, and the modern Scottish Gaelic literary tradition continues to this day, supported by a small but active community of writers, scholars, and readers.
The Pictish Problem
Pictish, the language of the Picts of northern and eastern Scotland before and during the early medieval period, is known only from a small and difficult body of evidence: a handful of inscriptions on stone, a small number of place names and personal names, and a few late medieval notes that may preserve older forms. The classification of Pictish is one of the most intractable problems in Celtic linguistics. The older view, that Pictish was a non-Indo-European language related to Basque, has now been largely abandoned. The current scholarly view is that Pictish is most likely a Celtic language, possibly a member of the Brythonic branch, but the evidence is not strong enough to allow a confident classification, and some scholars have proposed that Pictish may have been a Q-Celtic language distinct from Goidelic or Brythonic. The inscriptions, the late Katherine Forsyth’s 1997 Studies in the Archaeology of the Picts and the more recent work of Gordon Noble and Nicholas Evans, are essential for any serious engagement with the question.
The Ogham Tradition
The Ogham script is one of the most distinctive features of the early Irish linguistic landscape. Developed in the early centuries CE, Ogham consists of a series of lines and notches cut along the edge of a stone, with each group of strokes representing a letter of the early Irish alphabet. The script is found on several hundred stones scattered across Ireland, with a smaller number found in Wales, Cornwall, Scotland, the Isle of Man, and even in Devon and Silchester in England. The great majority of Ogham stones are memorial inscriptions, and they preserve a substantial body of early Irish personal names and genealogical information.
The origins and purpose of Ogham are still debated, but it is generally thought to have been devised in southern Ireland in the first few centuries CE, possibly as a response to contact with the Latin script used by Roman missionaries and administrators. The use of Ogham appears to have been largely confined to the inscription of names and genealogies on stone, and the script was never used for extended literary or administrative texts. Nonetheless, the Ogham tradition is an invaluable source for our knowledge of early Irish phonology, and it gives us one of the clearest windows into the linguistic state of the Celtic languages at the very beginning of their written record.
The Manuscript Tradition of the Celtic World
The medieval Celtic literary tradition is preserved in a remarkable series of manuscripts, most of which were produced in monastic scriptoria in Ireland, Wales, Scotland, and Brittany. These manuscripts, often beautifully illuminated, are the work of scribes who combined the older native traditions of craftsmanship with the new Christian learning. The Book of Kells, the Book of Durrow, the Lindisfarne Gospels, the Book of the Dun Cow, the Book of Leinster, the Red Book of Hergest, and many other great manuscripts are the vehicles by which the older oral tradition of the Celtic world was preserved for posterity.
The manuscript tradition of the Celtic world is particularly important because it preserves a substantial amount of material that would otherwise have been lost. The great mythological cycles of Irish tradition, the heroic tales of the Ulster cycle and the Fenian cycle, the legendary history preserved in the Lebor Gabála Érenn (the Book of Invasions), the Welsh Mabinogion, and the Breton lays of Marie de France all depend on the manuscript tradition for their survival. Without the patient work of generations of monastic scribes, much of the older Celtic literary heritage would be lost to us forever.
Sources and Further Reading
David Stifter’s Old Irish (2018) and Ranko Matasović’s A Concise Historical Grammar of Old Irish (2021) are the most useful single introductions to the oldest insular Celtic language. For the comparative linguistics of the family as a whole, Kim McCone’s Towards a Relative Chronology of Ancient and Medieval Celtic Sound Change (Maynooth, 1996) and the essays in Martin J. Ball and Nicole Müller (eds.), The Celtic Languages (Routledge, 2nd ed. 2009) are the standard references. For Ogham, Damian McManus’s A Guide to Ogam (1991) and R. A. S. Macalister’s Corpus Inscriptionum Insularum Celticarum (4 vols., 1945) are indispensable. For the medieval literary tradition, Robin Flower, The Irish Tradition (1947, repr. 1994), Patrick K. Ford, The Mabinogi and Other Medieval Welsh Tales (1977), and Joseph Clancy, The Earliest Welsh Poetry (1971) are the standard introductions. For Pictish, the late Katherine Forsyth’s Studies in the Archaeology of the Picts (Edinburgh, 1997) and the more recent work of Gordon Noble and Nicholas Evans are essential.